When President Trump named Oklahoma’s Scott Pruitt to head the Environmental Protection Agency (EPA) in 2016, he was a controversial pick. But Pruitt’s policy positions quickly became overshadowed by another set of problems when it was revealed that he crossed various ethical lines while in office. In one instance, he rented a condo co-owned by the wife of an energy lobbyist at a below-market rate. In another, he hired a dubiously massive security detail. In the end, it was these scandals that forced his resignation, and a big part of the reason these scandals ever came to light was that journalists on the EPA beat used the Freedom of Information Act (FOIA) to obtain records that helped them write their exposés.

That’s hardly the only recent example of FOIA making a difference. One journalist broke the story, using FOIA, that established the political origins of the Justice Department’s 2017 request that the census include a citizenship question (rather than the stated law enforcement justification). Yet another reporter used FOIA to document a pattern of negligent medical care in immigrant-only private prisons run under a federal government contract. While no single news story typically provokes immediate change, these pieces were among the reasons why Pruitt’s tenure at the EPA came to an end, the citizenship question was not included on the census, and the federal government announced in 2016 that it would let its contracts with privately run prisons expire.

Unfortunately, though, while these examples demonstrate FOIA’s potential, they represent bright spots amidst a sea of frustration and failure. By and large, journalists’ loudest complaints about FOIA are not that its guarantees of access to information are inadequate, but that agencies are not complying with them. This noncompliance is not sporadic; it is endemic. Despite hundreds of dedicated personnel working in FOIA offices across the federal government who believe in transparency, compliance failure is baked into FOIA’s structure because of incentives that systematically cut against releasing information to the public.

When the Freedom of Information Act was enacted in 1966, it was revolutionary. Its basic premise—that ordinary citizens have a right to know what the government knows—radically reimagined the relationship between the public and the federal government. Through FOIA, “people are permitted to know what their government is up to,” the Supreme Court explained in 1989, quoting historian Henry Steele Commager. To accomplish this lofty goal, Congress adopted a simple model: Anyone can request any government record for any reason, and the government must hand it over within 20 business days, unless the record falls within one of a limited number of exemptions to disclosure meant to protect interests such as privacy, national security, and trade secrets. The press was at the helm of the decade-long advocacy effort that led to FOIA’s passage, and the news media championed a vision of access to public records under which the public would use the law to hold government to account and to further our democratic ideals.

At times, this vision has come into spectacular reality. Journalists have used FOIA to uncover government waste and corruption, learn what kinds of outside influence impact government decision-making, and peek behind the curtain of the national security state. But more than a half-century after Congress passed this landmark legislation, it is clear that FOIA largely has not—and cannot—live up to its mission.

The core problem is this: There is no arm of the U.S. government that champions transparency. No government agency embodies a transparency mission; no court possesses transparency expertise. FOIA provides a core right for the public to access government records, but it lacks a locus for implementation and enforcement. Offices that administer FOIA are nested within every government agency but are often seen as detracting from—or at least distracting from—the agency’s core mission, whether that mission is protecting the environment, securing our borders, or guaranteeing the integrity of our economic marketplaces. Because they are not centrally organized, these agency FOIA offices can’t bring their collective wisdom to bear on FOIA implementation. With no mission-driven body at the helm to poke and prod and pester and inspire agencies to see their transparency obligations as being of central importance to their work and to the public, FOIA has atrophied. We need a transparency guardian.

Problem Number One: Delay, Delay, Delay

The biggest barrier that FOIA requesters encounter is also the most mundane: delay. Agencies shamelessly violate the 20-business-day deadline to respond to a request, often by months and sometimes by years. Backlogs, or the queue of requests that are overdue, run in the hundreds of thousands, and reporters complain of receiving stale information, oftentimes well beyond when the relevant story was published. Of course, sometimes agencies never respond at all—a particularly egregious form of delay. In the last reported fiscal year, even for requests designated as “simple track,” the average processing time across all agencies of government was nearly twice the 20-day limit. One ProPublica reporter waited three and a half years—as he put it, almost enough to put a child through college—just for his request to be summarily denied. At the end of FY 2023, the Air Force had a pending request that dated back to 2013; the State Department had multiple requests that dated to 2007.

It is hard to imagine that a statutory mandate can be so routinely and flagrantly violated. And yet, these delays are all but inevitable. The main reason is that FOIA operations are vastly under-resourced, and no mechanism exists to incentivize greater investment. Congress does not typically appropriate funds specifically for FOIA administration, and thus agencies must fund their FOIA programs from general-use dollars. As such, resources dedicated to FOIA directly compete with other agency priorities, leaving FOIA offices chronically underfunded. Meanwhile, the number of FOIA requests is enormous, now topping one million a year across the federal government. Agency FOIA offices report that they simply cannot keep up. For example, at Immigration and Customs Enforcement (ICE), a FOIA office with a comparatively large staff, last year 47 dedicated FOIA personnel received some 45,000 requests, alongside their other FOIA administration duties such as processing appeals, maintaining required online disclosures, supervising, reporting, and more. Perhaps it should be no surprise, then, that at the end of that year, ICE reported a backlog of more than 28,000 requests.

Agencies are capable of complying with FOIA when they shift priorities and commit the necessary resources. There have been occasions when courts have intervened and mandated that agencies clear backlogs and observe deadlines, and in those cases even some of the busiest FOIA offices, like U.S. Citizenship and Immigration Services, have met their legal obligations thanks to the additional resources they received from their respective agencies. But those are rare exceptions.

The challenge that underfunded FOIA offices face is magnified by the necessary involvement of non-FOIA personnel to fulfill most requests. For example, when immigrants’ rights organizations wanted more information about how agencies were administering a newly announced Migrant Protection Protocol (otherwise known as the “remain in Mexico” policy) in 2019, they filed a FOIA request for guidance documents, instructions, protocols, and communications between top officials related to the program. But those records were necessarily located in those officials’ inboxes, computer drives, and paper records. In such cases, officials (or their staff) have to search for those records and provide input on any context that would counsel toward redaction or withholding. Thus, significant intra- (and sometimes inter-) agency consultation is required. At this stage, again, FOIA competes with other agency priorities as personnel working on agency programs must take time away to assist the FOIA office.

On top of its demands on people’s time, transparency is uncomfortable and, at times, frankly annoying to agencies. Outside of FOIA offices, few government workers chose to enter public service because they’re keen on handing over information to curious requesters. Officials widely report that agency staff whose primary job duty isn’t FOIA view transparency as the kind of administrative compliance function that takes a back seat to other, more mission-driven work. As a result, even when FOIA offices themselves improve their processing speeds, they are hindered by foot-dragging throughout the agency. And while the FOIA offices may see areas ripe for improvement or reforms that could help alleviate delays, they often have little power within the agency structure to make those kinds of systemic changes. Delays stemming from underfunded FOIA offices are therefore compounded by the delays caused by other offices within the agency.

Other Problems: Exemptions and Redactions

Unfortunately, noncompliance goes far beyond delay. The heart of FOIA obligations—to turn over all nonexempt records upon request—is another area of systemic failure. Agencies routinely over-claim exemptions in order to withhold records that should be made public or to over-redact records to conceal more information than is justified. To take one prominent example: Classified information is exempt from disclosure, but experts in and outside of government agree that overclassification is a rampant and unchecked problem throughout the national security state. The criteria for classification, set out in a series of executive orders dating back to 1951, are ambiguous enough that they sweep in (at last count, in 2017) some 60,000 documents that contain new secrets a year, with another 50 million so-called “derivatively” classified documents that incorporate previously classified secrets.

In a poignant example, the Bush Administration released in May 2008 a redacted version of the 2004 CIA Inspector General Report on Torture. Just over a year later, the Obama Administration released a much more lightly redacted version. The side-by-side comparison reveals the sometimes arbitrary nature of classification decisions. It is unclear why, for example, the older version obscures a part of the table of contents that we now know describes a subset of the “Discussion” portion of the report as containing the subheadings “DCI Confinement Guidelines,” “DCI Interrogation Guidelines,” “Medical Guidelines,” and “Training for Interrogations.” How can these subheadings alone be sensitive enough to be classified?

Another exemption, called the deliberative process privilege, is designed to protect internal agency deliberations so that government officials can have frank discussions among themselves. But it is so often abused that it has been dubbed the “withhold it because you want to” rule. Congress, recognizing the problem, attempted to curb the use of this exemption with a 2016 FOIA amendment that banned its invocation for records more than 25 years old. Still, this is of little comfort to journalists trying to learn about government activities today.

Even more confounding for requesters, agencies may issue a “no records” response, asserting that there are no records responsive to the request, even when the requester has good reason to believe the records plainly must exist. To be sure, agencies are required to conduct only “reasonable” searches for records, not perfect searches, but it often seems dubious as to whether such efforts were made. For example, when Immigration and Customs Enforcement announced in 2022 that it had grown its “alternatives to detention” program of using ankle monitors in lieu of incarceration by over 20,000 individuals in just two weeks, one government data-tracking research organization filed a request for underlying source documents for that statistic. Yet, ICE responded with a “no records” response, claiming it had no case-level records whatsoever of individuals entering this program.

Over-withholding and over-redaction arise because, again, the incentives tilt against compliance at every turn. Where there is any plausible argument that a record—or a portion of it—is exempt from disclosure, a FOIA officer has every reason to err on the side of caution and claim the exemption. If the officer releases the record and it does result in some sort of harm, whether to a private interest, the agency’s reputation, or (perhaps worst of all) to national security, there are likely to be repercussions for the officer. By contrast, erring on the side of withholding the record never gets you in trouble. While a provision of FOIA authorizes sanctions of individual employees in cases where a court finds egregious violations of the act, that provision has literally never been invoked. And positive incentives are completely absent: Seasoned FOIA litigator Alan Morrison has asked, rhetorically, “Has anyone ever been awarded a medal or received a promotion for releasing any government records?” No, as far as any of us can tell. With disclosure a matter of high risk and no reward, the result is all but inevitable. A former head of the agency that oversees the classification system testified in 2016 that a “when in doubt, classify” attitude prevails within government, and the same could be said of withholding records generally.

In some even more troubling instances, agencies withhold records not because of unconscious risk aversion, but as a result of overt agency protectionism or political interference. For example, agencies have come under fire for implementing what are known as “awareness review” policies. These are procedures that require senior employees or their staff to review and revise certain proposed responses to FOIA requests before they are issued to the requester. At the Department of the Interior, one such policy led to a report in 2022 revealing that awareness review had been used to delay the release of records about then-nominee for Secretary of the Interior David Bernhardt, protecting him from criticism ahead of his confirmation hearing. A system with so much agency discretion is ripe for this type of abuse without the check of meaningful oversight.

Judicial Review Has Been Calamitous

The root of the compliance failure can be directly traced to the way Congress chose to have transparency obligations administered and enforced. In short, FOIA administration is left largely up to agencies themselves—a dispersed set of hundreds of separate entities all creating their own implementation schemes—and policed at the outer edges only by those FOIA disputes that make it to the federal courts. Federal court, however, is a last resort. At the outset, when an agency fails to meet a deadline, denies access to records, or redacts information, the requester can administratively appeal that denial of access. But that appeal goes to another official within the very same agency, not to an outside or independent body or adjudicator. In most cases, only if an appeal is denied is further recourse available by filing a lawsuit in federal court to challenge the agency’s action.

The legislative history reveals that Congress viewed the courts as the right locus for enforcement of the statute precisely because the courts are independent of the executive branch on which FOIA’s obligations are imposed. As such, FOIA specifies a particularly searching standard of review, called de novo review, under which judges are to make a completely independent decision about whether the agency has an obligation to grant public access to certain records with no deference owed to the agency’s initial determination. The congressional record at the time of FOIA’s passage in the 1960s shows that lawmakers believed the de novo standard was essential to “prevent [judicial review] from becoming meaningless judicial sanctioning of agency discretion.”

The theory of locating FOIA’s enforcement power in the judiciary may have been rational, but the practice has been calamitous. First, it takes little imagination to realize the time, money, and expertise needed to litigate in federal court. Most requesters will not have ready access to a lawyer, and few will be able to go it alone unrepresented. Plus, few lawyers take such cases pro bono since few requesters win these cases, and even fewer win them decisively enough to recover attorney’s fees from the government. Requesters are further discouraged from pursuing this remedy by the fact that litigation in federal court is extremely slow. If delay was the initial problem—as it is for more and more FOIA plaintiffs—a years-long court battle is hardly an effective remedy.

Pitifully few would-be FOIA plaintiffs ever challenge the agency decisions denying them access to federal records. In fiscal year 2023, for example, there were 1,122,211 requests processed in the federal government, only 16 percent of which were granted in full, meaning that the vast majority of requesters received a response other than access to all the records they wanted. But that same year, the federal government received only 14,443 administrative appeals, representing only around 1.2 percent of decisions under FOIA. And judicial review was all the more scarce; in recent years, the total number of lawsuits filed in federal court challenging agency FOIA decisions has averaged between 700 and 800 per year, meaning that less than 0.1 percent of potential FOIA litigants make it into court.

What does this mean for agencies? Consciously or unconsciously, agency officials know that the likelihood of having to defend any individual FOIA denial in court is vanishingly small. Judicial review that materializes in a miniscule number of cases provides no meaningful incentive—individual or structural—for agencies to improve compliance.

The effect of judicial review is further undermined by the underwhelming results of litigation even for those few plaintiffs who make it to court. Despite the de novo standard of review that Congress mandates, federal courts routinely defer to agencies’ representations of key underlying facts and determinations. One empirical study from 2002 found that agencies prevailed in FOIA cases at a far higher rate than in other types of agency review, even though in FOIA cases agencies are supposed to receive less deference, not more.

It is true that FOIA’s exemptions are drafted in broad strokes, and their application to particular records requires a certain amount of interpretation. Time and time again, however, the courts have interpreted the statute in a way that defers to agency secrecy. For example, as to an exemption that applies to certain “records compiled for law enforcement purposes,” some circuit courts have crafted a presumption that all records held by agencies whose primary purpose is law enforcement meet this threshold requirement. In another example, some circuits describe their review in national security cases as subject to a “substantial weight” standard, essentially replacing the independent review standard that Congress chose with a more deferential standard borrowed from courts’ stance toward affidavits in which agencies raise national security concerns.

Courts are also using a variety of procedural shortcuts that favor agencies. For example, judges nearly always refuse to exercise their power to examine the underlying disputed records themselves, known as in camera review, which prevents true scrutiny of agencies’ claims of exemption. They also deny access to discovery tools traditionally used in litigation, which prevents plaintiffs from interrogating the agency’s justifications for withholding. And agencies often are allowed multiple bites at the apple to try to meet their burden of proof. Judges occasionally express their disdain for FOIA cases, plainly finding them burdensome and difficult to adjudicate. As a result, courts fall victim to the same kinds of institutionally biased decision-making that agencies do: a bias that favors secrecy.

But just as agency FOIA offices shouldn’t be lampooned, neither should judges be blamed for the structural problems they face. FOIA cases often involve huge volumes of records, and judges lack the necessary expertise to understand their contents and context. Creating shortcuts and deferring to agencies—whether expressly or implicitly—is all but inevitable. And the federal courts can only decide those cases that are brought, which are neither representative nor plentiful. In short, the judiciary is a poor vehicle for policing FOIA obligations.

A Model for Reform

The United States was the third country to adopt an access-to-information (ATI) law, but it set a standard that provided a template for reforms around the world. Today, there are 140 countries with ATI laws, the centerpieces of which are typically request-and-response provisions much akin to FOIA’s. But these laws have also diverged from and, in important respects, innovated far beyond FOIA. Authors of newer laws have recognized that having a right to information on paper does little good without compliance and have thus created independent bodies—typically styled as information commissions—dedicated to implementation and enforcement. Indeed, 90 of the laws worldwide provide for an appeal to an independent administrative body, rather than relying on courts. The independent commissions in 52 of those countries, including Australia, Canada, Mexico, India, and the United Kingdom, can issue binding determinations about the public’s right to access the requested records, rather than merely request release through advisory decisions.

With the support of a Fulbright research grant, I spent the 2022-2023 academic year in Mexico City studying Mexico’s information commission, known as INAI for its initials in Spanish. INAI, which was created in 2002, is the world’s most celebrated independent information commission, blessed with strong constitutional protections that promote its independence and a full range of implementation and enforcement powers. When I arrived in Mexico, my goal was to study INAI as a model of agency adjudication—an alternative to the U.S. process of judicial review. That is, I wanted to understand the possible advantages and disadvantages of having an administrative body, rather than a court, resolve complaints brought by the public about possible transparency violations. By the time I left, though, I had come to understand how agency adjudication fit into a much larger and more holistic set of activities designed to protect Mexican democracy. In the end, I identified five categories of work that contribute to INAI’s successful implementation and enforcement of Mexico’s transparency laws. Versions of these five features could be adopted into a well-designed independent agency tasked with enforcing FOIA in the United States.

The first aspect of INAI’s work that I found so critical is the National Transparency Platform, and it touches every part of the transparency system. An online portal designed and maintained by an in-house team of computer engineers, the platform is the Swiss Army knife of transparency: It is a tool that can be used by each person for their own important ends. For a requester, it is a one-stop shop for submitting requests to any government body at any level of government, tracking all their requests in one place, receiving all their responses, and, with the click of a button, appealing any response with which they are dissatisfied to INAI. For INAI, the platform maintains a full record of everything that happened before a case gets to the commission for review, keeps track of all deadlines, facilitates communication with the responding agency, and allows INAI to pull data on compliance in an instant. For agencies, it acts as an internal tracking mechanism and facilitates delegation of tasks to appropriate units. And for the general public, it acts as a repository of all records required to be made affirmatively available without a prior request as well as all requests and responses issued to others. The platform, known as the PNT, comes up in almost every conversation about transparency in Mexico, and its importance cannot be overstated.

The second bucket of work that is central to INAI’s success is the part of its operation I was initially so focused on as an alternative to judicial review: adjudication of transparency disputes. INAI is an adjudication machine. It resolves around 20,000 disputes a year, and it does so extremely quickly, typically within 40 business days. Everyone I spoke with emphasized that requesters do not need expertise or a lawyer to engage the process. In fact, INAI staff repeatedly stressed that INAI has a duty to “supplement the complaint,” which is to say, to figure out what requesters should have argued if only they better understood government processes and legal requirements. In the adjudication process, INAI puts itself in the shoes of the requester. As one commissioner told me, “If you ask me, ‘Do you feel impartial?’ I will tell you, ‘No, I am partial. I am partial in favor of society.’” The then-president of the commission similarly explained, “INAI is allied with journalists.” With a constitutionally protected independent commission on their side, the requesters win, at least in part, more than three-quarters of the time.

The third bucket of work I will loosely term “monitoring.” This function includes various ways in which INAI monitors compliance with the law other than simply adjudicating the disputes brought to it. For starters, INAI has an enforcement unit specifically to make sure the government complies with the commission’s orders, and it is armed with the power to sanction noncompliance by various methods up to and including direct monetary sanctions against the government official responsible for the violation. A separate unit of the commission monitors compliance with the law’s obligations to affirmatively publish listed categories of records without a predicate request and grades agencies on their performance on an annual basis. To promote uniformity, INAI samples responses to requests that weren’t appealed in search of patterns of over-withholding or noncompliance. This is possible only because the PNT is a comprehensive repository of all requests government-wide. And not all of the commission’s power comes from its ability to bludgeon recalcitrant agencies: It also issues awards to agencies that do a particularly good job, which agencies themselves proudly displayed to me when I visited.

The fourth bucket could be termed “outreach,” or perhaps “education.” INAI engages the public regularly through every mechanism it can find. It purchases ad space in newspapers and on buses, hosts booths at book festivals and community events, staffs a public-facing office people can walk into in Mexico City, and operates a phone line where a human will help callers find the information they want. In a country where over seven million people speak an indigenous language, INAI translates materials into multiple indigenous languages. And to ensure that transparency becomes embedded in Mexicans’ sense of how their government should operate, the commission publishes children’s books and school curricula about transparency. This kind of promotion means that ordinary Mexicans are more aware of transparency’s role in their democracy and talk about it as a basic human right. Indeed, basically every Uber driver I spoke with knew of and had opinions about INAI—something I can assure you does not happen in the United States when someone asks me what my research is about and I say “FOIA.”

Finally, the last bucket is the plenum, the seven-member body that heads the commission. They hold public sessions most weeks that are broadcast on INAI’s website, during which they highlight key cases that demonstrate the importance of transparency in government. In addition, the commissioners represent the institution before the Mexican Congress and in public, and, symbolically, they set a broader tone for discourse around the role of transparency in democracy.

INAI is not perfect, and it has continually sought to innovate and improve. But the basic successes of its core functions have been celebrated by civil society, utilized by journalists, and verified by academics. It has been viewed as an essential check on abuse of government power, a necessary component in a democracy, and an institution that brings a citizen’s voice into government. Public satisfaction among citizens who engage the transparency system is shockingly high—upwards of 80 percent—and in my series of interviews with journalists who use the law, I found broad agreement about its efficacy, despite there being areas for improvement.

Yet, just recently, INAI has fallen victim to the politics of the day. Indeed one marker of INAI’s strength and importance in checking government power is the set of attacks on the institution mounted by then-President Andrés Manuel López Obrador (often referred to by his initials, AMLO), who decried INAI as “useless” and declared the guarantee of access to information a “farce.” But these attacks came after INAI created significant hurdles for AMLO’s domestic agenda. For example, when AMLO issued a decree in 2021 that certain infrastructure projects—including his signature Tren Maya, a train in the Yucatán peninsula—were matters of national security and thus shielded from various forms of legal scrutiny, INAI sued to establish that the projects should be subject to transparency obligations and won in the Supreme Court.

AMLO’s agenda carried over into the new Claudia Sheinbaum administration, which in short order moved to push through constitutional reforms that will result in the elimination of INAI and absorption of its functions into other federal agencies. President Sheinbaum has contended that INAI—and other independent oversight bodies like it—are expensive, unnecessary, and sometimes themselves corrupt. But the origins of the attacks in the AMLO administration suggest that INAI is worth attacking at least in part because it has done what it’s supposed to do: It has helped Mexicans find out what their government is doing that government officials wish the public didn’t know.

Time for a Transparency Revolution

Almost 60 years after FOIA was enacted, the United States needs a transparency guardian. An independent commission charged with promoting transparency would revolutionize FOIA in three fundamental ways, all leading toward genuine government compliance with existing obligations.

At present, there is nothing like INAI in the United States. The Office of Information Policy at the Department of Justice is tasked with issuing guidance to agencies on FOIA implementation, and the Office of Government Information Services, part of the National Archives, can offer voluntary mediation services for requesters and agencies. Neither is an independent agency with the protections that status entails, nor are they authorized to issue binding orders resolving disputes or otherwise actually enforce the statutory provisions of FOIA.

The first sea change a transparency guardian would accomplish would be to increase the sheer amount of oversight. Agency adjudication can be nimbler and involve less procedural formality than judicial review, providing more access to lay people and more efficient disposition of cases. Without the need for litigation, it would be much easier for the requester to challenge a denial of access to information. A transparency guardian could investigate all potential violations without regard to how well they are pled by the requester and would routinely examine the disputed records and question agency officials whenever there is doubt. Moreover, consolidating technology platforms and transparency infrastructure would alleviate burdens on FOIA offices in agencies and create efficiencies of scale.

Second, a transparency guardian would be mission-oriented and possess the expertise needed to protect the public’s interest in access to information. Just as the EPA attracts people who take seriously the difficult tasks of protecting the environment, and the National Labor Relations Board hires people who are not afraid to get into the weeds of labor disputes, a transparency guardian would attract personnel who believe that FOIA is essential to our democratic ideals and that the public’s right to know must be vigilantly guarded. The result would be much more careful, much more purposeful oversight that is likely to better account for the importance of the public’s interest when weighing close cases.

Perhaps most importantly, a transparency guardian can do many things a court can never do. It can use its powers to audit, gather data, and investigate to uncover systemic problems. It can educate and empower; it can advise and cajole; it can support and champion transparency inside of government and out. Any court is limited to resolving the disputes brought before it. But a transparency guardian can set its own affirmative agenda, seeking to improve transparency systematically rather than one case at a time.

These facets of a transparency guardian would go a long way toward addressing FOIA’s fundamental failures. Improved and increased oversight would change the incentives for agencies, ensuring they dedicate more resources and push against the natural inclination toward excessive secrecy. Dedicated funding from Congress for the transparency guardian itself could centralize and streamline many of the administrative operations that agencies currently must devise individually. And finally, a transparency guardian could change the culture around FOIA, helping public servants and the citizenry alike understand its centrality to the project of democratic governance.

Every so often, Congress takes up amendments to FOIA, almost always pushing in the direction of strengthening the public’s right to access information. But those reforms have largely nipped around the edges of the problems plaguing requesters and agencies alike. The time for structural reform has come. We must bring FOIA up to modern standards. Congress should give us, the public, a transparency guardian.