Drawing of man carrying papers through sludge

Abner Dean

Drawing from Abner Dean’s What Am I Doing Here? (1947), reissued by New York Review Comics

The Office of Management and Budget is required by law to produce a widely neglected annual report, the Information Collection Budget of the United States Government (ICB), which quantifies the annual paperwork burden that the government imposes on its citizens. The most recent ICB finds that in 2015, Americans spent 9.78 billion hours on federal paperwork.1

The Treasury Department, including the Internal Revenue Service, accounted for the vast majority of the total: 7.36 billion hours. The Department of Health and Human Services was responsible for 696 million hours imposed on (among others) doctors, hospitals, and the beneficiaries of Medicare, Medicaid, and the Affordable Care Act. The Department of Transportation accounted for no less than 214 million hours, including elaborate requirements imposed on truck drivers, automobile companies, railroads, and airlines. Comparatively speaking, the 91 million annual hours that came from the Department of Education might not seem like much, but for administrators, teachers, and students, they were pretty burdensome.

The ICB does not make for riveting reading, but it is worth pausing over those 9.78 billion hours. Suppose we insisted that for the entirety of 2019 all 2.7 million citizens of Chicago must work forty hours a week at a single task: filling out federal forms. By the end of 2019, they would not have come within four billion hours of the 2015 total. The Paperwork Reduction Act (PRA) was enacted in 1980 in an effort to reduce this burden, but it doesn’t appear to be living up to its name. (Disclosure, or perhaps confession: from 2009 to 2012, I served as the administrator of the Office of Information and Regulatory Affairs [OIRA], and in that capacity I oversaw administration of the PRA.)

If we use the average value for an hour of work as calculated by the Bureau of Labor Statistics—$27—then those 9.78 billion hours are the equivalent of $264.06 billion. That is more than double the budget of the Department of State and the Department of Transportation, and more than triple the budget of the Department of Education. And the monetary figures greatly understate the problem. Paperwork burdens can make it difficult or impossible for people to enjoy fundamental rights, such as the right to vote or to obtain life-changing benefits—or to avoid crushing hardships.

For paperwork burdens, the University of Chicago economist and Nobel laureate Richard Thaler has coined a good term: “sludge.” You might want to sign your child up for free school meals, but wading through the sludge might defeat you. To get financial aid for college, students have to fill out the Free Application for Federal Student Aid (FAFSA). It’s long and complicated; many students give up and fail to apply to college at all. The right to vote may be the most fundamental right of all, but if the registration process is full of sludge, a lot of people might end up disenfranchised.

Reducing these burdens can have an enormous impact on people’s lives. For free school meals, the Department of Agriculture has adopted a “direct certification” program, which means that if the locality or the school district has enough information to know that children are eligible, they are automatically enrolled. In recent years, more than 11 million children have benefited from the program (about 91 percent of the eligible population). Simplification of the FAFSA dramatically increases the likelihood that low-income students will apply for aid and eventually enroll in college. A number of states have adopted automatic voter registration, which means that if eligible citizens interact with a state agency (for example, by receiving a driver’s license), they are registered as voters. In 2016, Oregon’s automatic registration program produced more than 250,000 new voters; almost 100,000 of them voted. The private sector can also do a lot more to reduce sludge—for example, it can help people choosing among health care plans by simplifying options and explaining what is likely to be their best choice.

Paperwork burdens are often a product of decisions by public officials. At the national level, those decisions come from Congress or from administrative agencies, which have diverse motivations. Many of their goals are legitimate. Perhaps most important, they might be trying to ensure that people are actually eligible for the benefits for which they are applying. In addition, officials might be trying to collect data that they can use to improve their programs. Sometimes they use paperwork burdens as a kind of rationing device, limiting licenses, permits, or money to people who are willing to run some kind of gauntlet.

OIRA’s main website, reginfo.gov, provides a relatively clear picture of the “information collection requests” (the technical term for paperwork burdens) that it is currently deciding whether to approve. At any given time, OIRA might be assessing about one thousand such requests. Most end up being approved—perhaps because the law requires them, perhaps because they are justified, perhaps because they are not sufficiently scrutinized.


In their new book, Pamela Herd and Donald Moynihan use the phrase “administrative burdens” to include not just paperwork requirements but also an assortment of other obstacles, including waiting times, fees, and in-person interviews. In their view, administrative burdens undermine individual rights, above all the right to vote and the right to choose to have an abortion. They are keenly interested in distributive questions—in the effects of administrative burdens on specific social groups (including the poor, the disabled, and the elderly). They note that administrative burdens can be reduced and sometimes even eliminated by legislators or administrators who conclude that they do not make much sense. For example, benefits can be given automatically, as noted above, and they can be made simple to obtain, as is the case, in many states, for drivers’ licenses. But officials often require people to navigate an obstacle course, as is the case for those seeking political asylum in the United States.

Herd and Moynihan argue that administrative burdens are responsible for three kinds of costs: learning costs, psychological costs, and compliance costs. To obtain financial aid, for example, high school students might have to learn a great deal about their parents’ financial situation. Psychological costs include frustration and possibly stigma and humiliation; welfare applicants frequently experience these. Compliance costs include the time spent waiting in line, filling out forms, hiring accountants or lawyers, or paying processing fees.

For many people these costs prove overwhelming. That helps explain the low participation rates for many important federal and state programs—and the immense difficulty that people often have in obtaining permits or licenses of various sorts. Behavioral economists emphasize that an assortment of human biases amplify the real-world effects of administrative burdens. For example, it is often tempting to put off administrative burdens until another day. That day may never come, even if the consequences of delay are quite serious.

Herd and Moynihan offer a series of case studies. In some of them, burdens are a serious problem, but in others, Congress and federal agencies have worked hard to ensure that they are negligible. Their star performer is Social Security, which is simple and in important respects automatic. As they put it, “the biggest bookkeeping organization in the world banished burdens.” The Social Security Administration (SSA) does almost everything and generally requires citizens to do very little. It tracks people’s earnings and determines eligibility and benefit levels—automatically. If you are eligible, you can enroll online or go to one of the nation’s 1,200 field offices. Waiting times are usually short. After you become entitled to benefits, you are likely to receive direct deposits into your bank account within a month.

In the 1930s, many people took the administrative challenge as a serious objection to the very idea of Social Security. Partly because President Roosevelt insisted on making the system work, and partly because a large public constituency demanded it, the federal government succeeded in meeting that challenge. It did so in part through the creation of Social Security numbers, which make it much simpler to track people’s earnings over their working lives. In the United States, the poverty rate among older adults is now just 9 percent; if Social Security were not included in their income, it would be 40 percent. Almost one third of beneficiaries rely on the program for at least 90 percent of their income.

Herd and Moynihan are also enthusiastic about the design of the Earned Income Tax Credit (EITC), by which the federal government gives low-income workers an economic subsidy in the form of a credit. Most programs designed to benefit poor people have participation rates of between 30 and 60 percent; for the EITC, it is about 80 percent. As a result, the EITC is among the most effective antipoverty programs. Because it makes work more remunerative, it increases people’s incentive to work. The EITC makes a major dent in the national poverty rate and helps children in particular, with beneficial effects on their health, cognitive abilities, and long-term educational prospects.

The relatively high participation rate is a result of minimal sludge. The paperwork requirements are modest—a standard tax return is all that is necessary—and the IRS sends simple, clear reminders to people who appear to be eligible; the reminders significantly increase participation rates. The IRS also runs voluntary programs that provide free tax help. Because participation involves little in the way of frustration or stigma, the psychological costs are low.

True, the EITC is not as simple or automatic as Social Security. If 20 percent of eligible people are not receiving it, there is a serious problem. The IRS almost certainly could enroll people automatically. Nonetheless, the administrative burdens for recipients are much lower than they might be. An important reason is an unlikely overlap of political interests between left and right. Businesses are enthusiastic about the EITC, because it helps them attract workers, and they are joined by antipoverty groups, above all those seeking to help the working poor. There is a lesson to take from this: administrative burdens are an emphatically political choice.


Roosevelt and Churchill
Roosevelt and Churchill; drawing by David Levine

In other cases, there are no such coalitions, which reduces officials’ incentive to cut administrative burdens. Indeed, burdens are sometimes created precisely in order to make access more difficult. As an example, Herd and Moynihan point to the Affordable Care Act. Twenty-seven states opposed to Obamacare refused to create exchanges, which were supposed to provide a simple way for people to buy health insurance. Citizens in those states have had to apply through federal exchanges. The process can take months, and it has left millions of people in limbo. The Obama administration tried to reduce burdens by publicizing the program, reminding eligible people of its availability, and giving them relatively clear instructions about how to apply, and also by simplifying the federal–state coordination process. But the Trump administration reduced funding for publicity and shortened the sign-up period.

Medicare is a nearly universal program aimed at older people. If you are eligible for Social Security, you are usually eligible for Medicare too. The sludge comes from the astonishingly complex process faced by Medicare enrollees who are choosing among services. What is the right supplemental insurance plan? What is the right prescription drug plan? Is a Medicare Advantage Plan a good idea? These can be difficult questions, and as Herd and Moynihan point out, they are especially challenging for older adults, who often suffer from cognitive decline. They quote a Medicare beneficiary who complains, “That’s what gets me, they wait until we retire to make it complicated.”

There is a lot of evidence that Medicare beneficiaries are making bad choices and losing a lot of money in the process. For such programs, Herd and Moynihan argue that the government should “reduce choice and simplify options” with the use of online tools, telephone assistance (with shorter waiting times), and customized recommendations. In many cases, federal agencies have the legal authority to adopt such approaches on their own. The real question is whether there is political will—and whether burden-reduction is sufficiently urgent to lead anyone to take action.

What about constitutional rights? Under current law, states are allowed to regulate the right to abortion as long as their regulations do not impose an “undue burden.” Many states have taken that as an invitation to discourage abortion.

For example, some states have required pregnant women to get counseling, which includes descriptions not only of the procedure but also of fetal pain. (Many experts say that fetuses can’t feel pain until the third trimester; most abortions take place in the first two trimesters.) They have required women to take and see an ultrasound, to make multiple visits to clinics, and to undergo significant waiting periods. These requirements impose obvious costs. Herd and Moynihan offer an extensive quotation from a thirty-five-year-old woman describing her experience of navigating administrative burdens in Wisconsin:

I am shaken. I am embarrassed. I am scared. I am tired of waiting. I am now called into a room. I can bring Hubby this time. We are told to watch a video, again required by state law. The video talks about adoption, foster parenthood, the dangers of abortion, my rights. It drags on. I feel like a small child. Husband looks concerned and helpless. I sign a form indicating my understanding of the information presented on the video. We wait. A nurse finally comes back in. Time to go back to the waiting room. We’ll call you in a short while.

According to one study, administrative burdens increase the cost of abortion by 19 percent and decrease the number of abortions by 13 to 15 percent. One consequence appears to be an increase in the incidence of self-administered abortions. Herd and Moynihan’s analysis here, treating abortion as part of a broad category of administrative burdens, is especially timely: with the confirmation of Brett Kavanaugh, there is every reason to think that the Supreme Court will be increasingly receptive to the imposition of new burdens on women seeking abortions.

The Fifteenth Amendment, forbidding denial of the vote “on account of race, color, or previous condition of servitude,” was ratified in 1870, but for decades, administrative burdens have been used to disenfranchise African-Americans. Literacy tests were a favorite instrument until they were finally forbidden by the Voting Rights Act of 1965. In recent years, administrative burdens have become less onerous; voting is more convenient and registration is generally easier. But obstacles continue to exist, and in some states, they are mounting. Worst of all, they are being used as a political weapon, above all by Republicans seeking to increase their prospects for winning elections.

Some states are purging people from the voter rolls if they have not voted for a specified number of years or responded to a notice. As a result they are required to register again. Others are purging voters on the basis of flawed technologies designed to prevent noncitizens or felons from voting. Some states require a state-issued photo ID. That might not seem so taxing, but according to some estimates, about 11 percent of Americans do not have one. States have increased residency requirements and mandated proof of citizenship. Administrative burdens of multiple kinds are working to disenfranchise African-Americans, the elderly, and people with low incomes.

There are major differences, of course, among all these cases. Social Security, the EITC, the Affordable Care Act, Medicaid, Medicare, abortion, and voting rights do not present the same problems. But Herd and Moynihan convincingly show that it is instructive to see administrative burdens as a single category, reducing and discouraging access in various ways—sometimes for better but often for worse.

Herd and Moynihan conclude with a broad plea for reform. They want federal legislators and administrators to focus on reducing burdens on people who seek access to programs, benefits, and legal rights, perhaps by simplifying requirements (as the Social Security program has done) and perhaps through the use of automatic enrollment (as the Department of Agriculture has done for school meals). As they stress, automatic enrollment can drive administrative burdens down to zero. Where automatic enrollment is not possible, officials might use an assortment of tools: frequent reminders; simple language; online, telephone, or in-person help; and welcoming messages to reduce the psychological cost of dealing with overwhelming bureaucracy.

The authors also ask for an evidence-based approach to administrative burdens, including an effort to weigh their benefits against their costs. For example, are they really helping to reduce fraud? Voting is a case in which an evidence-based approach to the assessment of such burdens would be helpful; the same is true of federal benefits programs, including the EITC and the Affordable Care Act. Herd and Moynihan offer a checklist that could be used by administrators who want to learn about participation rates and how they vary across populations; the magnitude of compliance costs, in terms of both time and money; and the ease or difficulty, for potential participants, of finding out about a program and learning whether they are eligible.

By putting a spotlight on the consequences of administrative burdens, Herd and Moynihan have done a great public service. In many cases, participation rates can be dramatically increased with a mere shift from requiring people to apply (“opt-in”) to automatically enrolling them (“opt-out”). The best-studied area involves savings. When employers shift to automatic enrollment, the number of employees who participate in pension programs increases dramatically. If the goal is to increase the use of clean energy sources, reducing greenhouse gases, automatic enrollment also has a major impact.

To be sure, automatic enrollment is not always a good idea. Agencies may lack the capacity to know who should be enrolled. With automatic enrollment, people may find themselves participating in, and perhaps paying for, a program they do not like. But in many cases, simplification and burden reduction do not merely reduce frustration; they greatly improve people’s lives.

An underlying reason is that our mental resources are limited, and inevitably we are able to focus on only a subset of life’s challenges. If people are busy, poor, disabled, or elderly, overcoming administrative burdens can be especially difficult.2 Herd and Moynihan deserve particular credit for focusing on the distributional effects of administrative burdens—on whom they are most likely to harm. It is no surprise that in their case studies, the answer is often the poorest. If you are poor, you have to attend to a wide range of immediately pressing problems. If the government asks you to navigate a complex system or to fill out a lot of forms, you might simply give up.3

It is true that in many cases, administrative burdens have a plausible justification. The most obvious go by the name of “program integrity.” Suppose that the Internal Revenue Service decided just to send the Earned Income Tax Credit to apparently eligible taxpayers. It is nearly inevitable 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 criteria. If so, officials must ask which is worse: a design that will lead to some eligible people not receiving a benefit or a design that will lead to some ineligible people receiving a benefit? The answer should depend at least in part on the size of the two groups, though some people are especially troubled if ineligible people end up receiving taxpayer-funded benefits.

Administrative burdens might also be designed to counteract self-control problems and impulsivity. Small administrative burdens are frequently imposed online, with questions asking whether you are “sure you want to” send an e-mail without a subject line, activate a ticket, cancel a recent order, or delete a file. Those burdens, usually coming from the private sector, can be extremely helpful in making people consider their decisions carefully. More broadly, a degree of sludge, imposed by public as well as private institutions, might make sense for life-altering decisions, such as marriage and divorce.

It is true that we might not always trust public officials—legislators and bureaucrats—to decide whether sludge is a good strategy for helping people to avoid reckless decisions. But it is hardly impossible, in light of high-stakes choices, to defend administrative burdens as an effort to promote reflection and to provide valuable information. The accessibility of abortion is a more controversial case, of course. People disagree about whether administrative burdens are a reasonable means of discouraging ill-informed or insufficiently considered abortion decisions, or whether they merely pressure women not to exercise their constitutional rights.

There is also a growing literature on “hassles” and “ordeals,” exploring how administrative burdens might operate as a rationing device, ensuring that certain goods go to those who most want or need them. In some cases, such burdens might not be an irrational way of screening applicants for job training or other programs. If people are really willing to face administrative burdens, we might have good reason to think that they will benefit from those programs. At the same time, it is usually better to find more accurate and speedier ways to identify the people who are most in need of public services.

Privacy is also important. Government agencies sometimes impose administrative burdens in order to obtain information about people’s backgrounds—their employment history, their criminal record, their income, their wealth, their places of residence. If such agencies reduced paperwork requirements or made the process simpler and more automatic, they might well be required to find ways to obtain such information on their own and without people’s consent. That might be perceived as invasive, even intolerable. In some cases, there is a trade-off between frustrating burdens on the one hand and potential invasions of privacy on the other.

Notwithstanding these concerns, Herd and Moynihan convincingly demonstrate that administrative burdens impose a high cost, and that that cost often falls on those who are in the worst position to bear it. Citizens aren’t yet marching on behalf of a Sludge Reduction Act, but it would do a great deal to improve their lives.