The Elusive Quest for Government “Success”

Peter Schuck’s recent book, Why Government Fails So Often: And How It Can Do Better, is a scholarly tour de force. With each page, Schuck delivers blow after devastating blow to any hope that government programs hold the solutions to society’s major domestic problems. He demonstrates clearly the when, how, and why of government failure. When? Almost always. How? Let me count the ways. Why? Don’t get me started. Both the author and his book are worthy of the extensive acclaim they have received. From the Wall Street Journal and the New York Times to the Daily Show, commentators have rightly praised Schuck’s Herculean effort to focus our attention on the simple, disheartening reality of government failure. The reader comes away with two overwhelmingly depressing realizations: first, government programs frequently fail; and second, the cultural and institutional obstacles to reform are monumental. Professor Schuck builds a solid case for the proposition in his book’s title (“government fails”). But tellingly, if the book has any weakness it is that his effort to convince readers of the proposition in its subtitle (“it can do better”) is more fluid. Indeed, by laying bare the reality that many popular social programs are “unmitigated disasters” (as he puts it), as well as painstakingly documenting the many obstacles to systemic reform, Professor Schuck arguably undercuts his own effort to pave a more effective path forward. Perhaps the biggest impediment to convincing readers that government can succeed is convincing them that, according to some fair and objective measure, it has succeeded in the past. Professor Schuck employs the traditional yardstick for government success – cost-benefit analysis. In other words, he asks whether a government program accomplishes its objectives in a cost-effective manner. Yet, as he acknowledges, demonstrating “success” in the governmental context is complicated by host of factors, including imperfect information, the risk of moral hazard (think flood insurance and “too big to fail”), the persistence of invisible victims of government policies (that is, people a given policy makes worse off), and the likelihood that government intervention will crowd out more efficient market providers. These factors call into question the effectiveness even of the handful of policies that Professor Schuck labels successes. Consider two examples. First, the Homestead Act successfully distributed federal land to tens of thousands of American settlers in the 1800s, but much of the allocated land was of poor quality and the parcel size unproductively small. As a result, as Michael Heller has observed, “people either stayed and starved or abandoned the land.” The promise of free land also encouraged thousands to seek to game the system by squatting on land that would eventually be distributed, including (in my own family’s case) in Indian Territory. Second, Social Security provides (in Schuck’s words) “an essential financial safety net beneath private savings” and “has largely eliminated poverty among the elderly.” But, as Schuck acknowledges, Social Security also discourages private retirement savings, distributes massive amounts of resources to relatively affluent recipients, and “succeeds” under economic demographic conditions that “no longer exist and are unlikely to return.” These difficulties are compounded with respect to the Medicaid program, which encourages the elderly to divest themselves of assets to qualify for long-term care insurance. In a market economy, the “crowding out” phenomenon is particularly problematic. Consider these more recent examples: Professor Schuck, to be clear, does not identify these other programs as examples of policy successes. On the contrary, he lambasts the federal student loan program as well as the early implementation of the Affordable Care Act. Yet as others laud these efforts, they illustrate all too well the way that the government’s good intentions can crowd out private efforts that might make more of a difference. A further complication is that policy effectiveness rarely can be measured scientifically. As Professor Schuck acknowledges, the gold standard of policy evaluation – randomized field trials – is frequently impossible in the governmental context. Occasionally limited resources or policy design necessitate the randomized assignment of government benefits, which in turn enable researchers to compare the welfare of participants and non-participants. For example, at my own university, colleagues in Notre Dame’s Lab for Economic Opportunity are conducting a randomized controlled study of the effectiveness of resource-intensive early childhood education vis-à-vis Head Start (a program that, as Professor Schuck observes, arguably fails). But, for reasons both practical and political, government benefits through programs like Head Start usually are not distributed randomly. It is difficult to imagine, for example, Social Security benefits, Affordable Care Act coverage, or federal student loans being allocated by lottery. As a result, even the best available program assessments leave open enough “what if” and “but for” questions to make one wonder if the government intervention in question might have made things worse. Why Government Fails fills in some of those gaps, but not in a way that alleviates nagging doubts about policy effectiveness. As a friendly critic has suggested, the book might well prove to be a “gateway drug” to libertarianism, especially for those (like myself) who already harbor skepticism about both the effectiveness of, and perverse incentives created by, many government programs. Professor Schuck, a self-professed “militant moderate” who voted twice for President Obama, forswears any such intention. But, as Gene Epstein has observed, Schuck’s masterful treatment of government failure suggests that the answer to the subtitle’s question may be that – more often than not – “government might do far better by doing far less.” This essay is part of The Regulatory Review’s seven-part series, Is Government Prone to Fail