Vermeule provides a reasoned analysis of the fundamental cases underlying Administrative Law-- in which he concludes that Dworkin's Empire does not apply and that the Courts have ceded authority out of the necessity to meet modern needs, to administrative agencies populated by non-lawyers .Provides a balanced view of Chevron Deference rather than the sky-is-falling and powers are being unseparated hysteria being balanced by the scope of restraint on those who would offend the Rule of Law. ( )

One of the most interesting books about the law I have read to date. The book is essentially both a nod and a poke in the eye to Dworkin in the area of administrative law. The book very consciously constructs a Dworkinian theory, using fit and justification to show that Dworkin's famous line about how judges are the princes of law's empire is untrue today, if it was ever true. The book's thesis is that for internal legal reasons, the law has abengated its authority to the administrative state, deferring on questions of fact and law. Overall the book is a fascinating combination of doctrine, theory and decision making theory. The book is strongly utilitarian and influenced by law & economics, so heretics beware.

The book argues that once the courts recognized that agencies deserved deference for its expertise on questions of facts, marginalist thinking about the added benefit and costs of judicial review and the inherent coherence seeking nature of the law pushed the law to deference to legal questions as well. Even on purely legal questions, such as the jurisdiction of the agency, the courts have deferred to the agency (Arlington v. FCC) for internally legal reasons. Once Chevron created judicial deference to the agency's interpretation in its governing statute, there was no reason to separate out some of the statute as jurisdictional questions reserved for the court. A similar reason applies to Auer deference, since Chenery II allows courts to choose between adjudication and rulemaking, the fear of agencies writing overly board rules and getting deferrence for its interpretation of those rules is incoherent; agencies already have the power to decide how much and when to bind itself. The book goes so far to discuss how current case law, while setting out the factors for due process in Mathews does not preclude judicial deference to agencies. The chapter on due process is weaker but honest in that it admits there are different strands of caselaw on judicial deference for due process. The general thesis is that the dominant line of cases is at least silent on who gets to decide due process. Vermeule argues that the agency as the politically accountable and expert on system design should decide on due process which involves accounting for the costs and benefits of due process. Vermeule also argues that the different forms of deference essentially shift power within agencies from lawyers to non-lawyers, since the law creates an acceptable range of answers instead of a judicially determined point.

The book has an interesting argument against the classic concern that administrative state is a violation of the separation of powers principle. Vermeule argues that the administrative state did not conquer or force the government to give it power. Rather the separate branches, after deliberation have decided that an entity with the functional powers of all three was needed to deal with governance. Vermeule makes the strong claim that even if the administrative state was dismantled, it would be created again because the circumstances leading to its creation would still exist. The book argues that the separation of powers curbs abuse, but curbing abuse is not the only virtue. The new dealers argued that combining the powers has its own benefits, which is that such an entity can increase its governance activity level, acquire information, and as concentrated entities balance private entities using common law to concentrate its own potentially abusive private power. The administrative state is the product of the classical branches realizing that these were goods that needed to be balanced against traditional fears of public abuse.

While I found the entire book to be interesting and a good review of doctrine, I found the chapters on decision making theory and judicial review the most novel and fascinating. The book argues that while the academy talks about hard look review, in reality few administrative actions have been struck down for that reason (there is some empirical work as well as the argument that the data is not skewed by strategic responses since there is some evidence that even at different levels of review the courts defer at all levels suggesting deference is the norm). Vermeule argues against any judicial imposition of rationality requirements that the agency consider each alternative, and pick the best one because there are many cases under uncertainty that agencies will not be able to justify its choice on first order reasons, instead relying on second order reasons. For example if there is Knightian uncertainty, the agency must take an arbitrary decision, or in some cases because of strategic interactions, a choice is better than no choice at all. Even in cases where the agency could invest to find information to quantify an uncertainty, it may be uncertain if that investment is worth the benefit (and so on). In those cases, there is no reason either to impose a primitive closure rule through the courts, since it will not be clear that the primitive rule will even lead to the results the court assumes they do. The book of course, cites plenty of case law that have recognized this principle of thin rationality, where the courts accept second order reasons for agency decisions. The courts additionally have no legal power nor should they impose quantifiable CBA. There are reasons why agencies would not make decisions by selecting the absolutely best alternative in a choice set. In some cases, the agency may use a satisficer rule if cost of information is high, and facts are uncertain, or if there is a mean/variance tradeoff, accuracy/speed tradeoff or asymmetric consequences. In those cases, the courts impose a straightjacket by insisting on optimization under first order principles. Instead of State Farm hard look, the dominant caselaw is and should be Baltimore Gas.

Though Vermeule claims that he makes no normative claims about the administrative state, the book is clearly in favor of the fourth branch of government. Whether or not one agrees with this view, this book is a must read for those with an interest in the subject. It makes bold provocative arguments, novel theories and combines it with an impressive breadth of knowledge. ( )