*This is the ninth post in a symposium on William Araiza’s Rebuilding Expertise: Creating Effective and Trustworthy Regulation in an Age of Doubt. All posts from this symposium can be found here.

I am pleased to participate in a symposium on Rebuilding Expertise, written by Professor William Araiza. My praise for this terrific book (included on its back cover) raves in part that Araiza’s contribution is “sweeping” and that it provides “colorful and keen insight into how the government has functioned under several presidents from Reagan through Trump.” I also note that this volume illustrates “that the increasing interest in deregulation from politicians and the American public alike stems from a fundamental lack of trust in the government.” 

An essential take-away from Araiza’s work is that the public’s growing distrust of career civil servants has resulted in a pervasive lack of appreciation for the positive contributions and sheer necessity of our administrative technocracy. According to Araiza, the steeply falling popularity of administrative expertise can be attributed to both actual attrition in its quality as well as an ongoing PR crisis. As to the former, Araiza shows the extent to which presidentialism has, in fact, damaged the development and application of administrative expertise. As to the latter, Araiza suggests that distrust is also caused by the public’s lack of understanding of the benefits of administrative expertise and exacerbated by politicians who manipulate popular perceptions of agencies to gather support for deregulatory aims. 

In this post, I take the critical perspective that the problem with administrative expertise goes beyond the fact that it has been excessively influenced by presidents, a point with which I otherwise agree. More specifically, institutional values endogenous to the administrative state may also warp expertise. In Administrative Subordination, forthcoming in The University of Chicago Law Review, I argue (among other things) that agencies’ emphasis on a competing public interest value, efficiency, impacts administrative expertise in ways that undercut both immigration and environmental justice. For this reason, the public, or at least particular segments of it, have valid reason to be skeptical of the enterprise of administrative expertise.

One justification for the administrative state is that it operates with greater efficiency than Congress does. Efficiency is key to maintaining the U.S. system of administration, given the vast requirements and pressures under which our federal agencies operate. And yet, agencies’ emphasis on efficiency can negatively influence the data collection and analysis that undergird expertise, which is ostensibly another facet of agencies that renders them more suitable to policymaking than Congress.

As to the acquisition of relevant information, agencies sometimes tread on marginalized communities by acting based on data that is relatively easy to gather but not as accurate as information that would be more onerous or costly to collect. I refer to this easier-to-obtain data as “information proxies.” Consider the following examples from national security and immigration enforcement, excerpted from my new article (citations omitted):

Characteristics of people are used as proxies for information. First, the implementation of anti-terrorism law by the Department of Homeland Security (DHS) targets undocumented immigrants as proxies for those potentially engaged in terrorist activity, in order to accomplish its goals. Second, Muslim identity serves as a proxy for terrorist threats as well. By drawing on the criteria of “noncitizen” and “Muslim” as proxies for connections to problematic international forces, agencies believe that they able to identify potential threats to security more quickly. But these forms of immigration and national security enforcement also lead to the targeting of and more severe punishments against ethnic and religious minorities. 

Another example of an information proxy is an arrest record. Arrest records are used to make decisions about who receives official immigration status and who is deported, even if the arrest did not lead to a conviction that would render a noncitizen ineligible for status. Agencies rely on arrest records because they are inexpensive and accessible; considered to be “close-enough” representations of agencies’ priorities; conserve funding allotted to immigration enforcement; and allow immigration agencies to coordinate and pool resources with criminal prosecutors and police, thus increasing the capacity of all the enforcement interests involved. However, immigration decisions that take arrests into consideration exacerbate the racialization of both immigration and criminal enforcement.

As to the efficient analysis of information, sometimes an agency’s need to manage the institutional burden required to effectively evaluate data leads to subpar outcomes for vulnerable people. Consider the following episode from environmental regulation (from the same work-in-progress, citations omitted): 

The Environmental Protection Agency (EPA) was tasked with carrying out a statute that has special provisions for safeguarding children from pesticides. Instead of focusing its risk assessment analysis on the most dangerous pesticides, the EPA chose instead to focus on pesticides that are no longer in use or that have low risk factors. The EPA’s failure to use the higher standard is motivated by the cost of the necessary “hundred-fold” factor test, as well as the agency’s difficulties obtaining the data it needs to engage in additional pesticide review. Even if the EPA could gather the amount of information required by the statute, it would have to engage in an intense analysis of each pesticide that would require several levels of administrative review. As a result of its decision to streamline its data analysis, the EPA allowed pesticide levels to remain at an unsafe level for children (as illustrated by a lawsuit and a report filed by the Natural Resources Defense Council).

Ultimately, efficiency norms and expertise interact in meaningful ways that may render the administration of national security, immigration, and environmental law harmful to marginalized communities.

In the aforementioned review of Rebuilding Expertise, I suggested that “Araiza’s prescription—that the bureaucracy be accorded expansive power and wide berth—is controversial, but may very well be the key to improving both the effectiveness and legitimacy of the administrative state.” Since then, I have ascertained that administrative discretion plays a concerning role in failures of expertise. 

Consider the examples discussed above. The use of personal identity as a proxy for “terrorist” is enabled by the intense discretion allotted to DHS in the application of national security law. Likewise, the decision to rely on arrest records in noncriminal, administrative settings, whether in a particular case or to set categorical priorities, is an exercise of administrative discretion by immigration agencies. In addition, the EPA made several discretionary choices to streamline the implementation of a complex and broad set of pesticide provisions, leading to harmful outcomes for children. 

To the extent an agency chooses to develop and apply expertise in a manner that subordinates values such as fairness and justice to the institutional benefits of efficiency, it may be that the exercise of administrative discretion requires more—not less—oversight by the political branches and constraint by intra-executive accountability structures. My work shows that this appears to be the case in regard to agencies (including and beyond DHS and the EPA) that engage in immigration enforcement and environmental regulation, to a degree that suggests administrative expertise is notably flawed in these contexts.

Bijal Shah is an Associate Professor of Law, Provost Faculty Fellow, and Dean’s Distinguished Scholar at Boston College Law School.