Peter Schuck's Why Government Fails So Often is one of the most important books of the year and may be one of the most important books of the decade. Although I have seen this prolific author's name over the years, I had never read any of his work. My loss. Fortunately, I have read every page--including endnotes--of his latest book, and it is a tour de force.
Schuck, the Simeon E. Baldwin Professor of Law Emeritus at Yale University, calls himself a "militant moderate." I'm not sure what the "militant" part refers to: my guess is that it's his word for "passionate." And, although he does come off as a political moderate, his reasoning is radical. That is, in virtually all of the many government policies and procedures he considers, Schuck goes to the root of the problem. He dissects the workings of government, explaining why it works so badly and creates so many problems. It isn't until Chapter 11 that he gets to his examples of policy successes, and most of them are either thin gruel or examples of successes resulting from reductions in government's reach. In his final chapter, Schuck advises everyone, including libertarians, "to accept" both the fact of government's many failures and his reasoning about those failures. But he then cautions conservatives to "accept the fact" that "big government is here to stay." Although Schuck accepts the permanence of big government, his analysis of government's failures is so well-argued, so fact-based, and so devastating that it made me want to ask him, why aren't you a libertarian?
To explain why government fails so often, Schuck must establish that it fails often. He does so throughout the book in a nice weaving of cause and effect. In Chapter Two, titled "Success, Failure, and In Between," he gives his measures of success and failure. Schuck has fairly demanding, but not unreasonable, standards for total success. To succeed, in his view, a policy should pass a cost-benefit test, be "fair," and be manageable. He then gives 14 more principles that a policy should comply with to be implemented. I won't state them all here, but all are reasonable. The first is that policymakers "should intervene only when it will correct a significant market failure." Another is that a program "should be target-efficient." Yet another is that cost/benefit analysis "should be used to retrospectively analyze the effectiveness of existing policies, not just proposed ones." The last of the 14 principles is that policymakers should avoid the "Nirvana fallacy." Correctly citing my mentor, University of California, Los Angeles economist Harold Demsetz, as the originator of that term, Schuck explains that the Nirvana fallacy is what one commits when viewing a policy choice "as if it were one between an ideal program and the existing, flawed one."