Notwithstanding what I have said thus far, sludge often serves important goals. Sometimes it is indispensable. It saves money and even lives. It combats some of the worst human impulses, including recklessness, cruelty, selfishness, and greed. We can readily imagine six possible justifications for sludge:
Each of these deserves extended discussion. None is going to get what it deserves. I will touch only on the essentials. As we will see, a great deal of sludge turns out to be justified, which means that the discussion in this chapter will provide a necessary sense of balance. In deciding whether to eliminate sludge, we need to know what sludge, exactly, we are talking about. Even so, the general point remains: there is far too much sludge out there.
When public officials impose sludge, it is often because of a desire to ensure that programs work in the way that the law requires. One reason involves eligibility restrictions. Sludge can ensure that those restrictions are respected. People should not receive Medicare, Medicaid, the EITC, or Social Security unless they are entitled to the relevant benefits, and sludge is often a way of collecting necessary information. Even in the context of voting rights, burdens of various sorts can be and often are justified as a means of ensuring that would-be voters meet existing legal requirements. For spending programs, a usual justification for sludge points to “fraud, waste, and abuse”;1 sludge can be an effort to reduce all three.
Sludge is often imposed to obtain information about people’s backgrounds—their employment history, their income, their criminal history (if any), their credit rating, their family history, their travel history, their places of residence. Those who seek to work in government, certainly at levels that involve national security, are required to provide a great deal of information of that sort.2 (It takes endless hours, and some people do not provide all of the required information.) Would-be employees do not enjoy all the sludge. Nonetheless, it might well be justified. When students and others face administrative burdens to get visas, or to otherwise visit the United States, it may be to ensure compliance with legal requirements and to protect national security.
What is true for the public sector is true for the private sector as well. Banks and other institutions generate a lot of sludge. Those who seek a loan face sludge. A central reason is to ensure that they actually qualify. If you want to get a mortgage, you might find the process time-consuming, annoying, endless, and a bit humiliating. But banks do not want to give a lot of money to people who cannot pay it back. It is true that many companies try to obtain customers by keeping sludge to a minimum. But firms depends on sludge to make sure that they are enrolling or helping the right people. When patients encounter sludge at hospitals, it might be for good reasons.
It is true that with the increasing availability of information and with machine learning, private and public institutions might be able to find the relevant information on their own. In the private sector, some companies use the idea of prequalification, which means that they have enough information to know, in advance, that some people are already qualified for goods or services.3 Sometimes forms can be prepopulated; as a result, forms might not be necessary.4 That is a blessing, and the government should take far more advantage of it. In the domain of taxation, a dramatic example is the idea of return-free filing, which eliminates the need for taxpayers to fill out forms at all.5 Some nations, such as Denmark, do that already. In the United States, reliance on return-free filing could save billions of hours in annual paperwork burdens. In the fullness of time, we should see significant movements in this direction.6
But those movements remain incipient. For the present and the near future, the most obvious justifications for sludge go by the name of program integrity.7 Suppose that the IRS decided to send the EITC to apparently eligible taxpayers. If it could do so at low cost, and if the apparently eligible taxpayers are in fact eligible, there would be little ground for objection. The problem, of course, is the word apparently. It is possible that some of the recipients will not in fact be eligible. Whenever people are automatically enrolled in a program, some of them may not meet the legal criteria.
When this is so, public officials must choose between (1) a design ensuring that some eligible people will not receive a benefit and (2) a design ensuring that some ineligible people will receive a benefit. That may not be an easy choice, and it may not even be easy to decide on the criteria that we should use to make it. If the idea of program integrity is meant to refer to the number of errors, the choice between (1) and (2) might turn purely on arithmetic: Which group is larger? If automatic enrollment, removing the sludge, means that 500,000 eligible people will receive the benefit who otherwise would not, and if a degree of sludge means that 499,999 ineligible people will receive the benefit who otherwise would not, automatic enrollment might seem justified.
But it would be possible to see things differently. Suppose that automatic enrollment gives benefits to 200,000 eligible people but also to 200,100 ineligible people. Some people might think that if the 200,100 people are nearly eligible—if they are relatively poor—it is not so terrible if they receive some economic help. But other people might insist that taxpayer money is accompanied by clear restrictions and argue that if it is given out in violation of those restrictions, a grievous wrong has been committed. On this view, even a modest breach of program integrity, to the advantage of those who are not eligible, is unacceptable.
The most extreme version of this view would be that a grant of benefits to a very large number of eligibles would not outweigh the grant of benefits to a very small number of ineligibles. In my view, the most extreme version is hard to defend, unless the law requires it: a grant of benefits to a hundred people who are almost (but not quite) eligible is a price worth paying in exchange for a grant of benefits to a million people who are in fact eligible. But the correct trade-off is not self-evident, and reasonable people might differ.
The example might seem a bit tedious, but public officials encounter the problem all the time, and so do companies, universities, nonprofits, homeless shelters, hospitals, and others. As sludge increases, the class of people who receive a benefit may well include fewer who are not entitled to it. That’s good. As sludge is reduced, there is an ever-growing risk that a benefit will go to people who should not be getting it. That’s bad.
In some cases, of course, there is no need for sludge. We can obtain program integrity without it. In the direct certification program for school lunches, the level of accuracy appears to be very high; few ineligible children are allowed to qualify. It is increasingly possible to make things automatic, and so to eliminate sludge, without conferring benefits on the ineligible. The only point is that in many cases, trade-offs are inevitable, and different people can make different judgments.
Consider the question of voter registration. It would be great if we could eliminate sludge and make every eligible person a potential voter. Perhaps we can do that; as we have seen, the (excellent) idea of automatic voter registration is getting increasing support. Nonetheless, sludge has been defended as a way of combating the risk of fraud and thus ensuring the integrity of the voting process.8 On imaginable assumptions, sludge reduction could have the benefit of ensuring that eligible people are allowed to vote while also having the cost of ensuring that the same is true of (some) ineligible voters. The size of the two categories surely matters.
The most interesting, and perhaps the most profound, argument on behalf of sludge insists that far from deploring it, we should celebrate it, as a well-tailored response to human fallibility. In brief: Sludge of diverse kinds might be designed to promote better decisions—to counteract self-control problems, recklessness, and impulsivity. Sludge protects people against their own errors. For that reason, sludge can easily be judged as a sensible cure for a behavioral problem.
Behavioral scientists sometimes contrast two families of cognitive operations in the human mind: System 1, which is rapid, intuitive, and emotional, and System 2, which is deliberative and reflective.9 Sludge can be a way to strengthen the hand of System 2. When System 1 is agitated, passionate, or on fire, sludge is a kind of yellow light, allowing System 2 to take control.
Consider the case of mundane decisions, for which small amounts of sludge are frequently imposed online. As we have seen, it is standard for people to be asked whether they are sure they want to send an email with nasty language or without a subject line; make a payment; cancel a recent order; change a payment method or address; or delete a file. Such burdens can be an excellent idea. People might be acting impulsively; they might not be paying enough attention. Significant sludge, imposed by private and public institutions, might also make sense for life-altering decisions, such as marriage and divorce.10 If people are deciding whether to end a marriage, “cooling-off periods” can be a blessing.11 When emotions are leading people to make rash decisions, a mandatory waiting time might be useful as a way of slowing things down and making sure a momentary impulse is not at work.
Some sludge also makes sense before the purchase of guns, partly as a way of promoting deliberation.12 In fact, evidence suggests that waiting period laws, imposing sludge and delaying the purchase of firearms by a few days, reduce gun homicides by roughly 17 percent.13 The seventeen states (including the District of Columbia) with waiting periods eliminate about 750 gun homicides per year as a result of that policy. Expanding the waiting period policy to all other US states would prevent an additional 910 gun homicides per year without restricting who can own a gun. For preventing gun violence, creative uses of sludge could do a great deal of good. In the context of abortion, we have seen that some people defend sludge as a way of increasing reflection and deliberation.
Sometimes private and public institutions ask for a lot of personal information. Many people think that if they are to receive that information, it must be with people’s explicit consent. A pervasive question is whether to ask people to face time-consuming administrative burdens or instead to intrude on their privacy. Perhaps it is not so terrible if the government, at least, chooses the former.
At one period, of course, officials had no real option. Even if they wanted to intrude on privacy, they could not do so because they lacked the means. Increasingly, however, public institutions (and private ones as well) have independent access to that information, or they might be able to obtain it with a little effort. As a result, they are in a position to reduce sludge. Return, as a simple example, to the direct certification program of the US Department of Agriculture.14 Officials know who is poor, and so they can directly certify people as such.
In countless other cases, available data can enable private or public institutions to announce, very simply, that certain people are eligible and on what terms. They might know who lived where, who owes what, who committed what crimes, who traveled where, who spoke with whom and when. They might be able to prepopulate forms. Government agencies might be able to share data with one another.15 Facebook, Google, Twitter, Instagram, YouTube, and the US government could share data. To that extent, sludge can be a thing of the past.
But would that be desirable? To put it gently: not necessarily. In many cases, there is a trade-off between irritating burdens on the one hand and potential invasions of privacy on the other. Consider, for example, the question of how much information credit card companies should acquire, and should be allowed to acquire, before offering cards to customers. We might welcome situations in which such companies can find what they need and simply send eligible people offers or even cards. Or maybe we would not welcome that at all. Whether we should do so depends in part on what information they have and whether it might be misused.
The stakes are probably higher if government has or acquires the relevant information. In the worst cases, we might fear that public officials will use that information to punish their enemies. In harder but still troubling cases, we might fear that public officials will use that information for purposes that have little or nothing to do with the reasons for which it was originally acquired. Within the US government, there is a category called personally identifiable information (PII), and the law imposes severe restrictions on its acquisition and use. The very existence of the category suggests that for government, the risks of information acquisition might be thought unacceptable—and hence we might tolerate some sludge. Consider in this light the option of automatic enrollment, which eliminates sludge. That option might well be possible only if we authorize, and are able to get comfortable with, a great deal of information-gathering by institutions that people might not trust.
In recent years, the question of acquisition of information by private companies has become sharply debated. Many people think that the risk of abuse is greatest with government. But with mounting concerns about “surveillance capitalism,”16 that might not be obvious. Perhaps social media companies, or other big tech companies, can use what they know about each of us to manipulate us to make choices that are in their economic interest. Perhaps they can treat us as puppets on a string. If so, we might want to limit their ability to acquire information and also their ability to use what they get. Concerns of this kind go far beyond the topic of sludge. But if the best way to reduce sludge is to let private institutions acquire personal information on their own, we might think, well, sludge isn’t so bad.
The question of security is closely related. Sometimes institutions ask for sludge not to annoy anyone, but to make sure that information receives the protection it deserves. To set up an online account, for example, people might be asked to provide, and might be willing to provide, sensitive information. It might involve their bank account or their credit card. People might be asked to answer questions about their address, their Social Security number, or their mother’s maiden name. Sludge might be designed to ensure against security violations. Answering these questions is not exactly fun, but they might be justified as a means of ensuring against some kind of breach. Compare two-factor authentication, which imposes sludge. (And just speaking for myself, it is really, really annoying.)
Ideally, of course, we would have some clarity about the benefits and costs of sludge that is designed to protect people’s security (a point to which I will return). But if costs and benefits are difficult to specify, it might make sense to have a rough-and-ready sense that a degree of not especially onerous sludge is desirable to prevent the worst-case scenarios.
Might sludge be a way of targeting? Might it be a way of ensuring that goods go to the people who most deserve or need them? Might it turn out to be an effective and even fair strategy, the best available option in many settings?
A growing literature on hassles and ordeals explores that possibility. It shows that sludge might operate as a rationing device, increasing the likelihood that benefits, including licenses and permits, will be allocated as they ought to be.17 The intuitive idea is that burdens can improve self-selection. Suppose that a movie or a concert is immensely popular. If so, people might be asked to stay on the telephone or to wait in line for a ridiculously long time. If that approach can be justified, it is because an investment of time, like an expenditure of money, helps measure how intensely people want things. In the same vein, seemingly onerous administrative burdens, and apparently pointless sludge, might be a reasonable way of screening applicants for job training or other programs. If people are really willing to run the gauntlet, we might have good reason to think that they will benefit from those programs.
No one denies the importance of finding good ways to screen those who seek access to scarce resources. In markets, the willingness to pay criterion provides the standard screen; it is meant to ensure that people will receive goods if and only if they are willing to pay for them. If you are willing to pay a certain amount of money for apples or oranges, and no more, we will have a measure of how much you really want apples or oranges. The willingness to pay measure has obvious advantages. It’s pretty simple; people use it every day. It’s a lot better than randomness. But it also has major problems. Willingness to pay is dependent on ability to pay, and if you are not willing to pay much for apples, or for health care, it might be because you are poor. For many benefits coming from the government, willingness to pay does not make a lot of sense. Would we want to say that people get money to help them cope with a pandemic depending on how much they are willing to pay for it?
Willingness to pay money is just one way to measure people’s needs or desires. Another way is willingness to pay in terms of time and effort (WTPT, for short). How much time are you willing to pay to get a driver’s license? To get financial aid for education? Because willingness to pay depends on ability to pay, you might think that the willingness to pay criterion discriminates against people without much money. WTPT does not have that particular defect.18 If anything, it discriminates against people without much time. There may or may not be a correlation between lacking money and lacking time. In any case, government (or the private sector) might choose to use WTPT as a way of targeting—as a way of ensuring that goods are allocated to people who really need and want them.19
It is not exactly crazy to think that sludge can be a reasonable way of targeting. Intuitively, that idea makes some sense. The problem is that if the goal really is to target benefits, imposing sludge is often a singularly crude method of doing so. In fact, it can be terrible. A complex, barely comprehensible form for receiving federal aid is not exactly a reliable way to ensure that people who need financial help actually get that help. If public officials want to ensure that people who are eligible for the EITC actually receive it, a degree of sludge is not the best sorting mechanism. Ordeals have their purposes, and sludge can be an ordeal. But it is a hazardous mechanism for targeting.
Actually, it is worse than that. In some cases, ordeals work in concert with the limitations faced by poor people, so as specifically to select out those with the highest need!20
On reflection, that should not be so surprising. If those with the highest need are also the least likely to wade through sludge—if they suffer from scarcity of multiple kinds—sludge will fail as a targeting mechanism. Suppose that a public institution wants to make mental health services available to those who most need them. Suppose too that sludge is a particular deterrent to those suffering from depression or anxiety. If so, sludge will filter out exactly those people who most need the services. The same problem can be found if sludge is designed to ensure that scarce resources go to the neediest among the poor. It may be that the neediest are in the worst position to navigate sludge.
The problem is pervasive. It highlights a central point here. Sludge should be assessed for its distributional effects. If it has especially adverse consequences for the most disadvantaged members of society, it should be reduced or eliminated.
To run a government, public officials need a great deal of information. Among other things, they need to know how programs are working. Public officials might impose sludge to acquire data that can be used for multiple purposes, and that might benefit the public a great deal.
For example, officials might want to know whether people who receive employment training or some kind of educational funding are actually benefiting. What do they do with that training or that funding? Sludge might be essential to obtain answers to that question. Or suppose that the government is trying to reduce the spread of an infectious disease, to promote highway development, to monitor hazardous waste management, to ensure that pilots are properly certified and that airplanes are properly maintained, or to see how food-safety programs are working.21 Those who receive information-collection requests might complain of sludge. But the relevant burdens might be justified as a means of ensuring acquisition of important or even indispensable knowledge.
A related point involves record-keeping and ongoing monitoring. Public officials need to obtain information about what is happening on the ground, and that can require sludge. In many countries, public officials gave out an immense amount of money to help individuals and institutions cope with the ongoing COVID-19 pandemic, and also to help researchers of multiple kinds. What was the money used for? Where exactly did it go? It is not easy to get answers to those questions without requiring sludge. For people who are subject to it, the process might be burdensome, but if we want to manage public programs, we might well need to impose that burden.
In some of these cases, of course, sludge might be an effort to ensure program integrity. But I am emphasizing a different point. Even if program integrity is already guaranteed, officials might seek information, and require people to provide it, in order to provide both short-term and long-term benefits. Importantly, that information might be made public and used by private and public sectors alike.22 In the modern era, acquisition of information might promote public and private accountability. It might save money. It might spur innovation. It might even save lives.
These are important justifications for sludge, and they are easy to overlook. But they should not be taken as a kind of blank check or as an open invitation for officials to impose significant administrative burdens. For any particular burden, a central question is whether the government is actually acquiring useful information. If public officials are asking people to file with paper rather than electronically, refusing to reuse information that they already have, declining to prepopulate forms, or requiring quarterly rather than annual reporting, they should face a burden of justification. In all of these cases, they might well run into difficulty in meeting that burden.
In the abstract, it is not possible to say whether sludge is justified as a means of generating useful or important information. Some cases will be easy; any such justification will not be credible. Other cases will also be easy; any such justification is self-evidently convincing. Still other cases will be hard; without investigating the details with care, we cannot know whether such a justification is sufficient. The only point is that the benefit of sludge might be found there.
1. See 6 U.S.C. § 795 (2012) (“The Administrator shall ensure that all programs within the Agency administering Federal disaster relief assistance develop and maintain proper internal management controls to prevent and detect fraud, waste, and abuse.”); Jerry L. Mashaw & Theodore R. Marmor, Conceptualizing, Estimating, and Reforming Fraud, Waste, and Abuse in Healthcare Spending, 11 Yale J. on Reg. 455 (1994); Julie K. Taitsman, Educating Physicians to Prevent Fraud, Waste, and Abuse, 364 New Eng. J. Med. 102, 102 (2011).
2. For an example, see US Office of Pers. Mgmt., Standard Form 86: Questionnaire for National Security Positions (2010), https://www.opm.gov/forms/pdf_fill/sf86-non508.pdf (https://perma.cc/KB9P-JJ8D).
3. For a prescient discussion, see generally Ekambaram Paleenswaran & Mohan Kumaraswamy, Recent Advances and Proposed Improvements in Contractor Prequalification Methodologies, 36 Building & Env’t 73 (2001).
4. Note the emphasis on prepopulation in Memorandum from Neomi Rao, Admin., OIRA, to Chief Information Officers 8 (Aug. 6, 2018), https://www.whitehouse.gov/wp-content/uploads/2018/08/Minimizing-Paperwork-and-Reporting-Burdens-Data-Call-for-the-2018-ICB.pdf (https://perma.cc/KF9L-N6NZ), hereinafter Memorandum from Neomi Rao (Aug. 6, 2018) (“Sometimes agencies collect data that are unchanged from prior applications; in such circumstances, they may be able to use, or to give people the option to use, pre-populated electronic forms.”).
5. See Austan Goolsbee, Brookings Inst., The “Simple Return”: Reducing America’s Tax Burden through Return-Free Filing 2 (2006), https://www.brookings.edu/wp-content/uploads/2016/06/200607goolsbee.pdf (https://perma.cc/C695-5YQL) (“For the millions of taxpayers who could use the Simple Return, however, filing a tax return would entail nothing more than checking the numbers, signing the return, and then either sending a check or getting a refund.”).
6. See Memorandum from Neomi Rao (Aug. 6, 2018), supra note 4, at 8 (“Also worth considering is whether, in some circumstances, to dispense with forms entirely and to rely on more automatic, generic, or direct approval of participation.”).
7. See Protecting Americans from Tax Hikes (PATH) Act, Pub. L. No. 114–113, 129 Stat. 2242 (2015) (referring to Title II as “Program Integrity” and specifically intending to reduce fraudulent and improper payments in the EITC and other programs); Leslie Book et al., Insights from Behavioral Economics Can Improve Administration of the EITC, 37 Va. Tax Rev. 177, 180 (2018) (noting that “program integrity” of the EITC was an important topic among employees of the IRS because 43 to 50 percent of all EITC returns are incorrect, with most errors benefitting claimants); Program Integrity, Ctrs. for Medicare & Medicaid Servs., https://www.medicaid.gov/medicaid/program-integrity/index.html (https://perma.cc/2ZMC-XTSH) (Medicaid Program Integrity); Reducing Improper Payments, Soc. Sec. Admin., https://www.ssa.gov/improperpayments (https://perma.cc/T8ZN-XA32) (Social Security programs).
8. See, for example, Husted v. A. Philip Randolph Inst., 138 S. Ct. 1833, 1848 (“The NVRA plainly reflects Congress’s judgment that the failure to send back the card, coupled with the failure to vote during the period covering the next two general federal elections, is significant evidence that the addressee has moved.”).
9. See Daniel Kahneman, Thinking, Fast and Slow 13–15 (2011).
10. See, for example, Fla. Stat. Ann. § 741.04 (2018) (making the effective date of marriage licenses three days after application unless both partners take a premarital education course); Mass. Ann. Laws ch. 208, § 21 (LexisNexis 2018) (allowing divorce to become absolute ninety days after the initial judgment).
11. See Pamaria Rekaiti & Roger Van den Bergh, Cooling-Off Periods in the Consumer Laws of the EC Member States: A Comparative Law and Economics Approach, 23 J. Consumer Pol’y 371, 397 (2000) (“Cooling-off periods are potential remedies for the problems of irrational behaviour, situational monopoly, and informational asymmetry.”); Dainn Wie & Hyoungjong Kim, Between Calm and Passion: The Cooling-Off Period and Divorce Decisions in Korea, 21 Feminist Econ. 187, 209 (2015) (“The cooling-off period has no significant impact on divorce rates when the cause of divorce is … dishonesty, abuse, or discord with other family members.… Couples reporting the cause of divorce as personality difference or financial distress responded to the cooling-off periods.”).
12. See, for example, Cal. Penal Code § 26815(a) (2018) (requiring a waiting period of ten days for all firearm purchases).
13. Michael Luca et al., Handgun Waiting Periods Reduce Gun Deaths, 114 Proc. Nat’l Acads. Sci. 12162 (2017).
14. See U.S. Dep’t of Agriculture, Direct Certification in the National School Lunch Program Report to Congress: State Implementation Progress, School Year 2014-2015 2 (2016) (“Direct certification typically involves matching SNAP, TANF, and FDPIR records against student enrollment lists, at either the State or the LEA level.”).
15. On some of the relevant trade-offs, see generally Memorandum from Jeffrey D. Zients, Dep. Dir. for Mgmt., & Cass R. Sunstein, Admin., OIRA, to the Heads of Executive Departments and Agencies (Nov. 3, 2010), https://obamawhitehouse.archives.gov/sites/default/files/omb/memoranda/2011/m11-02.pdf (https://perma.cc/56QK-7HCR) (encouraging federal agencies to share data to improve program implementation while complying with privacy laws).
16. See Shoshana Zuboff, The Age of Surveillance Capitalism: The Fight for a Human Future at the New Frontier of Power (2019).
17. Examples include Albert Nichols & Richard Zeckhauser, Targeting Transfers through Restrictions on Recipients, 72 Am. Econ. Rev. 372 (1982); Vivi Alatas et al., Ordeal Mechanisms in Targeting: Theory and Evidence from a Field Experiment in Indonesia (NBER, Working Paper No. 19121, 2013), https://www.nber.org/papers/w19127.pdf (https://perma.cc/6XFF-QP8E); Amedeo Fossati & Rosella Levaggi, Public Expenditure Determination in a Mixed Market for Health Care (May 4, 2004) (unpublished manuscript), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=539382 (https://perma.cc/GF5A-YRY5); Sarika Gupta, Perils of the Paperwork: The Impact of Information and Application Assistance on Welfare Program Take-Up in India (Nov. 15, 2017) (unpublished Ph.D. job market paper, Harvard University Kennedy School of Government), https://scholar.harvard.edu/files/sarikagupta/files/gupta_jmp_11_1.pdf (https://perma.cc/K4HY-3YK4).
18. Note that if people are willing to pay others to do a relevant task, such as waiting in line or preparing taxes, the difference between WTP and WTPT might be erased.
19. The IRS provides free online tax preparation to 60 percent of taxpayers. IRS’ Intent to Enter into an Agreement with Free File Alliance, LLC (i.e., Free File Alliance), 67 Fed. Reg. 67,247 (Nov. 4, 2002). The program is available for free to taxpayers with income less than $66,000 annually. IRS, About the Free File Program (Nov. 21, 2018), https://www.irs.gov/e-file-providers/about-the-free-file-program (https://perma.cc/L5CL-X4ZG).
20. See Gupta, supra note 17, at 30–31.
21. Some of these examples are drawn from OIRA’s Information Collection Dashboard. Information Collection Review Dashboard, OIRA, https://www.reginfo.gov/public/jsp/PRA/praDashboard.myjsp?agency_cd=0000&agency_nm=All&reviewType=EX&from_page=index.jsp&sub_index=1 (https://perma.cc/8X7M-9RHE). For those who are interested in sludge reduction or in information collection in general, the dashboard (typically neglected by scholars) is worth careful attention.
22. See Data.gov, where the US government provides a great deal of useful information, much of it emerging from information-collection requests.