The administrative state is the central and unavoidable topic of modern constitutional theorizing. The single most striking difference between the constitutional arrangements of the late 18th century and today – “constitutional” in the sense of the actually obtaining structures and practices of government -- is that the modern state is, by any conceivable measure, largely an administrative state. In the institutional landscape of the late 18th century and well into the 19th, one may see precursors of the administrative state, but one simply does not observe the massive and elaborately reticulated bureaucracy that structures and constitutes the experience of government for almost all citizens.
The phenomenon is so vast and variegated that it is hard even to know how to approach it. Before we can explain or evaluate the administrative state, we need an organizing framework to bring some sort of intellectual order to the problems. In what follows I offer three such frameworks, respectively grounded in constitutional law, in democratic theory, and in institutional epistemology. The three are not mutually exclusive, nor do I argue that one is superior to another. Each of the three interacts with the others and, indeed, is partly constitutive of the others, as I will try to show. The final test of each framework is its utility, its fertility for producing substantive insights. What I do claim is that each framework makes sense of debates that actually occur among students of the administrative state, and supplies those debates with a regulated field on which the intellectual melee can occur.
The first framework involves the constitutionality of the administrative state (Part I), the second its democratic credentials (Part II), the third its epistemic capacity (Part III). Each implicates a set of debates with somewhat different, although overlapping, sociological substrates. The debate over constitutionality takes place principally among academic lawyers; the debate over democratic credentials takes place in a mixed community of academic lawyers, political theorists and political scientists; the debate over the epistemic capacity of the administrative state takes place among lawyers, economists, political economists, and policy analysts.
After describing each framework, I will ask how each affects the others (Part IV). Here I suggest that constitutionalists, democratic theorists, and epistemic technocrats – acting in the best of faith -- each see the concerns of the others as a source of unavoidable side constraints to be respected, while seeing their own concerns as a source of positive goals to be pursued. As a logical matter, not all of these perspectives can be institutionalized at any given time, although any one or two can be. The result is an endlessly shifting series of alliances, akin to a game of “divide-the-dollar” among three players – a game which has no stable core, no equilibrium solution. Law, democracy and knowledge will always contend for mastery of the administrative state.He concludes that
The administrative state is a site or terrain of endless tension, interaction and conflict among the three frameworks and their proponents. Any item in the triad may and routinely does come into tension or conflict with both of the other. Constitutionalists argue that even the claims of democratic accountability and expertise cannot trump the rule of law. Democrats argue that both experts and lawyers should be “on tap, not on top”; ultimately, citizens must decide through ordinary democratic processes what the scope and structure of the administrative state should be. Civil servants, technocrats, experts on advisory committees, and other knowledge workers tend to see both political accountability and legal constraints as distractions from, or interference with, their welfare-maximizing projects.
Tension and conflict are only part of the picture. As illustrated throughout, each framework tries to subsume the others, at least in part. Constitutionalists suggest that democratic accountability is itself a constitutional value; proponents of the “unitary executive,” who are hostile to independent agencies, are especially likely to argue that accountability running from the agencies, through the President, to the electorate is a sort of master-value of the Constitution. Democrats argue that democracy itself requires that law be available to constrain agencies and their experts to follow the commands of legislative representatives. Technocrats want optimal democracy and optimal legality, insofar as democracy and law themselves help to promote the welfare of the relevant population. And so on around the circle. The point is not that the proponents of one framework ignore the claims of the others; rather it is that those claims are taken into account, even in the best of faith, in a way that makes conflict inevitable at the level of the overall system. Each framework treats its own master principle as the criterion by which the claims of the others are assessed. To the constitutionalists, for example, democratic accountability is a value just insofar as the Constitution makes it so, and not an inch farther.
In practice, furthermore, each of these camps tends to see the claims of the other camps as side constraints to be complied with at the lowest possible cost, while seeing their own concerns as goals to be pursued to the maximum possible extent. It is not that policy experts, for example, think that law and democracy are worthless; what they tend to think is that social welfare (somehow defined) should be pursued unless there is some indisputable legal or democratic override. Likewise, democrats abstractly respect the role of constitutionalism and the claims of expertise, both of which are indispensable components of a well-functioning democratic order, but they see those values as collateral obligations to be satisfied while in pursuit of the main goal, that of democratic participation and accountability (somehow defined). And likewise for lawyers, who tend to see full compliance with constitutional norms as the central aim of their work, although they are willing to allow a sort of override to legality if compliance with law would produce disastrous policy outcomes or threaten serious harm to democracy. The override is triggered only once a high threshold of harm is reached. However sharp the disagreements about what law requires, it is the rare lawyer who cheerfully acknowledges that legal violations by the institutions of the administrative state, in pursuit of other goals, should be seen as acceptable whenever the overall balance of reasons so warrants.
Given scarcity of time and other resources, it seems unlikely in the extreme that the administrative state could be simultaneously perfectly democratic, perfectly informed, and perfectly lawful. And it is not merely unlikely, but indeed logically impossible, that more than one of the triad of law, democracy and knowledge could simultaneously be the sole maximand or goal of the system. If the proponents of each framework hold the structure of beliefs and motivations that I have claimed, then it is inevitable that all but one of the camps is condemned to disappointment at any given time. The observable result is an endless series of shifting alliances, in which legalists sometimes side with experts against “political interference”; experts and democrats sometimes join forces to curtail the claims of lawyers and judges; and democrats and lawyers sometimes ally to curb the pretensions of independent technocrats and experts. The administrative state resembles a divide-the-dollar game with three players, in which any two players may agree to divide the dollar between themselves, freezing out the third. The problem is that the players can reach no resolution that is stable in the long run, because a coalition of any two players can be broken by an improved offer from the losing party. Like the divide-the-dollar game with three players, the administrative state has no core, no equilibrium. It is a terrain on which law, democracy and knowledge endlessly assert their respective claims.