Corporate directors and other institutional decision-makers should add Cynical Theories to their reading list.
When They're Too Good at Their Job . . .
The 25th anniversary of James Q. Wilson’s Bureaucracy: What Government Agencies Do and Why They Do It marks an appropriate occasion to reflect on the contributions of this work to our understanding of bureaucratic behavior and performance, and the extensive—and, at least in some areas, growing—presence of the administrative state in the lives of American citizens. Jeremy Rabkin’s essay does both admirably. Rabkin lays out the complex picture of bureaucratic behavior painted by Wilson, and teases out the insights that might help us understanding recent high profile instances of apparent bureaucratic incompetence or misconduct.
I would like to use this short essay to highlight a dimension of the administrative state that receives little attention in Bureaucracy but seems to me particularly central for those concerned with the possibility of constitutionally constrained, democratic governance: The increasing scope for independent bureaucratic action that is largely beyond meaningful accountability to democratically-elected policymakers. This development is especially acute with respect to the national security state, as recent revelations regarding the extensive surveillance activities of the National Security Agency have highlighted. But they are not limited to that realm.
Ironically, this growing power of the bureaucracy is not a result of the inefficiencies of bureaucratic behavior highlighted by Wilson, but rather a product of the fact that administrative agencies are often very good at their job. More importantly, it poses an increasingly significant challenge to a constitutional order intended to guarantee the fundamental liberties of individuals, and to preserve the sovereignty of “We, the People.”
To develop this argument, it is useful to return to Wilson’s book and Rabkin’s application of it to current conditions. Perhaps the most notable (and, at times, frustrating) feature of Wilson’s account of bureaucratic behavior is that he offers no simple, parsimonious explanation. For Wilson, bureaucratic motivations are complex.
Bureaucrats are not simple budget maximizers (though they care about budgets), they do not always aim to maximize their department’s authority and jurisdiction (though they pay attention to it), nor do they pursue only personal, private benefits (though they care about these, too). Bureaucrats are shaped by past experience, motivated by peer groups, and molded by the organizational culture of their department. They are influenced by the tasks and goals of their organization, especially if these can be tied to readily observable outcomes. Moreover—and most important for current purposes—bureaucrats are responsive to the demands and constraints placed upon them by outside forces, including their “political masters”: the President, congressional majorities, committees, and the courts.
A recurrent theme in Wilson is that much of what is bemoaned about bureaucratic performance—rigidity, inflexibility, an apparent inability to move expeditiously in achieving particular results—is a result of the fact that bureaucrats are subject to multiple masters. Instead of following clear directives, they must navigate an environment characterized by competing and often contradictory and inconsistent demands.
Note a critical feature of this argument: Bureaucrats are, at some level, “victims” of the divisions within, and competing claims and constraints placed upon them by, the larger political system. Thus, for Wilson, ultimate blame for bureaucratic failure often belongs to the broader environment within which agencies must work.
Implicit in such a characterization is the supposition that if political stakeholders were to sing from the same hymn sheet, and provide clear direction to an agency, they could effectively manage bureaucratic performance. Put differently, in Wilson’s account, bureaucrats are ultimately agents of their political principals. They may not be particularly effective, and they may frustrate the ends of their masters, but they are clearly subordinate.
A similar picture emerges from Rabkin’s essay. Rather than highlight (as Wilson does) bureaucratic inefficiency, Rabkin focuses on the incompetence and misconduct of administrative agencies, especially as exemplified by recent scandals. But as in Wilson’s account, the central problem for Rabkin is that agencies are not particularly good at their job—and the challenge in institutional reform is how to make agencies more effective and efficient.
While these accounts obviously capture important elements of bureaucratic behavior, they fail to emphasize an important development in the modern administrative state: Rather than being ineffective and inefficient, contemporary government agencies are often too effective in the sense that much of what they do is beyond meaningful oversight by, and accountability to, democratically elected policymakers. Agencies have carved out large spaces for independent action, and are in the proverbial “driver’s seat” in determining significant portions of public policy.
Of course, this is not exclusively the doing of the bureaucracy. Congress and the President have found it convenient to devolve policy discretion to administrative agencies, and it should come as no surprise that these actors take such discretion “out for a spin,” to stick with the analogy. Nevertheless, this growing power of the bureaucracy poses a fundamental challenge to the ideals of constitutionally constrained, democratic governance.
Consider recent revelations regarding the expansive scope of domestic and international surveillance by American national security organizations, including the NSA and CIA. Whatever one may think of Edward Snowden, it is difficult to deny that his actions have exposed governmental surveillance of an extent that would have been unimaginable to most Americans only a few years ago. Moreover, members of Congress—including key members of the Senate Intelligence Committee—were not much better informed, and appear to have been left in the dark with respect to significant portions of NSA and CIA conduct.
Of course, by their nature, national security policy and the precise details of operations require some secrecy. At the same time, Congress obviously cannot exercise meaningful oversight of programs of which it is not aware. In addition—and this is a point to which I return below—the secrecy surrounding these programs prevents the kind of public debate that is critical in shaping public opinion, which in turn provides an important inducement for congressional action. Put differently, the motivation for elected policymakers to exercise meaningful oversight is lessened dramatically when agency actions are hidden from view, and not subject to public debate or scrutiny.
Perhaps more importantly, even if elected officials were fully aware of the scope of agency activities, there are significant hurdles to effective oversight. Virtually all public policy—including in the realm of national security—involves difficult trade-offs among competing values. With respect to national security, perhaps the central question concerns the extent to which personal freedoms and privacy should be restricted in order to increase the ability of government to monitor and potentially avert terrorist activity.
There is, obviously, no clear answer to this question, and individual citizens and policymakers will vary widely in their views of the matter. However—and this is a key point—how these competing values should be balanced in anyone’s mind often depends critically on information, including information about the nature of threats, relevant technology, available policy options, and the costs and effectiveness of these options.
Typically, such information must be provided by the very agencies involved. That is, in reviewing and overseeing the activities of agencies like the CIA and NSA, members of Congress and the President must often rely on information that is provided by agencies, and that may be difficult to assess independently. This informational asymmetry affords agencies a tremendous advantage. How other policymakers view and evaluate the options that are presented to them can be affected powerfully by how information is presented (or whether it is presented at all). This implies that agencies are often in a position to shape how those in a position to review and oversee their activities evaluate the inherent trade-offs involved in making policy.
Judicial oversight of the activities of national security agencies provides a poignant example. To engage in surveillance, these agencies must typically apply for a warrant from the Foreign Intelligence Surveillance Court, a court that meets in secret and whose opinions are not published. There is no adversarial process—only government lawyers appear before the judges. Of the 32,093 warrant requests the FISA court received from its creation in 1979 to 2012, only 11 were rejected. This is an extraordinary success rate that is surely explained—at least in part—by the fact that judges must rely on a one-sided stream of information supplied precisely by the agencies they are supposed to oversee.
The point is more general: The fact that critical information regarding the evaluation of agency activities must come from the agencies themselves makes it exceedingly difficult for democratically-accountable policymakers in the executive and legislative branches to exercise effective supervision. The process by which policy choices are made can become unhitched from democratic control.
Why does this matter for the ideal of constitutionally constrained, democratic governance? A bedrock principle for such governance is that given the trade-offs among competing values inherent in making public policy, those who decide on such trade-offs should ultimately be accountable to citizens. For legislators and the President, this principle is applied directly through fair and competitive elections. In the case of bureaucratic decision-makers, it is indirect, and bureaucrats must be accountable to a popularly chosen Congress and President. This chain of delegation is broken when agencies are beyond effective congressional and presidential oversight.
Moreover, there are good reasons, identified by Wilson in Bureaucracy, to suspect that those who work in national security organizations are likely to arrive at judgments regarding the appropriate balance among competing values in this area that differ from those that might arise out of a sustained, public debate. As Wilson highlights, bureaucrats are, in part, motivated by the mission of their agency. Individuals who are charged with protecting national security are likely to place high value on preventing attacks—this is, after all, the raison d’être for their organization. Furthermore, failure to prevent a terrorist attack is potentially disastrous for the agency, and the career prospects of its personnel. In contrast, the costs imposed by measures designed to reduce the likelihood of a terrorist attack will typically figure less prominently in the bureaucratic calculus. These costs are largely borne by others.
Besides, if the agency succeeds in preventing an attack, such success serves as immediate vindication for its policy choices. It is difficult for critics to demonstrate that a particular policy was not, in fact, critical to protecting the homeland.
As a result of these dynamics, those who work in these agencies are not likely to take a critical view of the expansion of governmental power at the expense of civil liberties. More importantly, the particular way in which they view this trade-off may color—whether deliberately or not—the manner in which they present policy choices and background information to Congress and the President. Coupled with the fact that most elected officials are wary of taking steps for which they might be blamed in the event of a terrorist attack, this “informational privilege” often leaves agencies in a strong position to guide policymaking, rather than being subject to oversight and direction by elected officials.
Let me emphasize that nothing I have said here implies that particular policies at issue are unwise or ill-considered. For example, a plausible argument can be made that widespread gathering of data on the electronic communications of ordinary citizens is desirable. This might be the case not because ordinary citizens are potential suspects, but precisely because they are not. Given the ocean of electronic communications around us, an effective strategy of monitoring and intercepting terrorist communications might require a baseline of what normal communications look like. Suspect communications only stick out when compared to the flood of what is ordinary. Given the threats involved, the security gains of such a policy might be so significant that most Americans would conclude that giving up some privacy, and accepting the potential abuses that will arise in a system in which such surveillance takes place, are justified.
However—and this is the critical point—in a democracy, how to strike this balance is a decision that should ultimately be made by policymakers accountable to citizens, preferably in the context of a vigorous public debate. These are not trade-offs that should be left to a bureaucracy—even if peopled by well-intentioned individuals—that is largely beyond effective accountability to the elected branches.
Gordon Tullock once observed that “having pointed out a major problem, it is the custom to suggest a solution. Unfortunately, in this case, I cannot do so.” I am afraid I find myself in a similar position. The complexity of public policy in advanced industrialized democracies increasingly affords administrative agencies an informational advantage vis-à-vis their political “masters.” This advantage is particularly pronounced in areas—such as national security—in which there is a dearth of reliable external sources of information (such as the interest groups or economic interests that provide an outside check in other policy arenas). Coupled with the fact that national security policy requires, for obvious reasons, a certain degree of secrecy, the challenge of ensuring democratic accountability of agency behavior is considerable. Nevertheless, it is a challenge that we must ultimately meet to preserve the ideal of constitutionally constrained, democratic governance.
 Gordon Tullock, “Public Decisions as Public Goods,” in Virginia Political Economy, edited by Charles K. Rowley (Indianapolis: Liberty Fund, 2004), p. 393.