1
Managing the Great Recession
A STRESS TEST FOR MODERN GOVERNANCE
Timothy J. Conlan, Paul L. Posner, and Priscilla M. Regan
ON FEBRUARY 17, 2009, President Barack Obama signed the American Recovery and Reinvestment Act (the Recovery Act, or ARRA) into law. This enormous piece of legislation was one of the most important legislative achievements of Obamaâs presidency, andâat $787 billionâit was by far the largest and most sophisticated piece of countercyclical, antirecession policy since the 1930s. In brief, the program provided an infusion of $275 billion in additional grant-in-aid funds to state and local governments, another $288 billion to citizens and corporations in the form of tax cuts, and a remaining $224 billion in a mix of federal spending for purposes such as enhanced information technology systems in health care, clean energy production and transmission, and improvements to federal facilities and infrastructure. Among the more than ninety different federal aid programs that received additional funds under the Recovery Act were a mix of highly flexible grants to the states; additional money for existing aid programs in education, social services, and infrastructure; and new grant programs designed to stimulate innovative programs at the state and local levels.
Though hotly debated in Congress before its enactment and widely discussed in the media during its first year, the Recovery Act has attracted less scholarly attention than warranted given the initiativeâs size, scope, and significance. To be sure, a number of economists have examined its macroeconomic effects and have produced varying, but mostly positive, estimates of its effects on economic growth and employment.1 A handful of other studies have sought to examine some elements of its impact on state and local finances or individual policy domains such as energy and education.2 But these efforts have left serious gaps in our understanding of this sweeping legislation, including several theoretical puzzles raised by ARRAâs implementation. This is most unfortunate. A central argument of this book is that the implementation of ARRA is uniquely informative about how large federal programs can be effectively implemented and held accountable in an era of indirect and networked governance, about how our political and governmental institutions respond to large doses of fiscal shock therapy, and about the wide discrepancy between public perceptions of governmental performance and realities on the ground.
For example, take the issue of disparate perceptions. A widely held public perception has been that ARRA was a colossal failure. A 2010 poll by the Pew Research Center found that nearly two-thirds of the responding public believed that the stimulus program failed to improve the nationâs employment situation, whereas only 33 percent believed that it helped.3 In contrast, 97 percent of economists polled by the Initiative on Global Markets at the University of Chicago believed that the stimulus initiative reduced unemployment. Moreover, two-thirds of economists believed that ARRAâs economic benefits outweighed the costs of funding the program, in taxes and borrowing, whereas only 5 percent believed the costs outweighed the benefits.4
What explains this discrepancy in public and expert assessments? No doubt it is due in part to economistsâ greater knowledge of macroeconomic theory in general and to specific estimates produced by economic models of the stimulus actâs effects. But the publicâs negative assessment of the Recovery Act was also a product of pervasive mistrust in the efficacy of government that has grown during the past several decades, fueled by unrelenting partisan attacks on the program and anecdotal media stories during and after its enactment. According to one survey, nearly half the public reported hearing mostly negative things about the stimulus program in news reporting.5 During the summer of 2009, as the first round of stimulus funds were being spent or obligated, examples of negative news stories proliferated. A $3.4 million project in Florida to create âturtle crossingsâ under a busy highway received considerable, though often inaccurate, attention in the media.6 So did improvements at a lightly used border crossing in Montana.7 Many of these stories emanated from a collection of so-called wasteful stimulus projects collected and publicized by Senator Tom Coburn (R-OK), an iconoclastic but very conservative skeptic about most forms of government spending.
Indeed, in a program of this size, widespread waste and complexity were universally expectedâby the media, which established crack investigation teams designed to root out the waste, which they assumed would be forthcoming;8 by public interest advocates, who encouraged citizen investigations designed to âcrowdsourceâ problems;9 and by the Obama administration itself, which feared negative political fallout from fraud and waste and thus built an unprecedented system of accountability and transparency into the program.10
However, there is surprisingly little evidence of widespread waste, fraud, and mismanagement in the Recovery Act programs. As a USA Today retrospective on the program put it, âThe financial losses under the 2009 Recovery Act have been just a fraction of what the government expected.â11 For example, in 2009 members of Congress were concerned that the Recovery Act might lose as much as $55 billion to waste and fraud, based on an estimate by the Association of Certified Fraud Examiners that 7 percent of funds are lost to fraud and waste in a typical program.12 Yet after five years of audits and investigations, auditors for the Recovery and Transparency Board found only $57 million in fraudulent spending, or just 0.007 percent of the total stimulus, despite ARRAâs unprecedented scope and pace of implementation.13 Consequently, ARRA is absent from Paul Lightâs inventory of forty-one government programs experiencing significant failures between 2001 and 2014.14
Indeed, many of the cases that initially seemed to be credible reports of waste or fraud in stimulus programs did not stand up on closer scrutiny, as in the case of the Florida turtle crossing project. In other cases, the Obama administration vigorously disputed the criticsâ characterizations of projects as wasteful or ineffective.15 Similarly, a Mercatus Center report in 2009, which claimed that the Obama administration was selectively steering stimulus funds to congressional districts represented by Democrats, received considerable media attention.16 As the political blogger Nate Silver observed, however, this report was fatally flawed. It failed to account for the fact that virtually all the congressional districts that received the most stimulus funds encompassed state capitals. As Silver noted, âA lot of stimulus funds are distributed to state agencies, which are then responsible for allocating and administering the funds to the presumed benefit of citizens throughout the state.â17 It just so happened that state capitals were more likely to be represented by Democrats than other, often less urbanized, areas. A subsequent, more sophisticated reanalysis of the data by two political scientists vindicated Silverâs critique. A paper by Jeffery Lazarus and Jason Reifler concluded that âthese findings indicate that stimulus funds are being distributed largely according to the criteria set out in the legislation. Once a House districtâs ability to substantively claim stimulus money is controlled for, its memberâs party is no longer a significant predictor of the level of funding it receives.â18
Theoretical Perspectives on the Recovery Actâs Implementation
Even if the public rhetoric surrounding the Recovery Act was misguided or exaggerated, serious puzzles remain for both scholars and public administrators. For example, the existing political science literature on policy implementation has traditionally been pessimistic about the performance of new, large-scale federal initiatives. This was certainly true of early studies by scholars such as Martha Derthick, Jeffrey Pressman, and Aaron Wildavsky. Their research efforts into the New-Town In-Town initiative of the late 1960s and the Economic Development Administrationâs urban areas grants were motivated by a desire to understand why both programs were colossal failures.19 Later research attempted to identify factors associated with the potential for successful implementation, while still conceding that âeffective implementation of major programs ⌠is exceedingly difficult.â20 And such problems remain. One can easily fast forward to contemporary research on more recent programs such as No Child Left Behind and find a comparable concern with why major federal programs fail to deliver on their promises.21
Moreover, the core findings of such implementation research make the design of the Recovery Act read like a virtual manual of what not to do. The literature consistently argues that factors such as program complexity, multiple layers of administration, rushed implementation, and ambiguous statutory instructions all tend to undermine effective program implementation. Yet the Recovery Act seemingly violated each and every one of these precepts.
For example, one context in which implementation research has found potential for successful implementation has been in programs that were given long periods of time to err, evolve, and improve. For example, this was a key finding of Michael Kirst and Richard Jungâs analysis of the Elementary and Secondary Edu...