This post is part of Notice & Comment’s symposium on Joshua D. Blank and Leigh Osofsky’s Automated Agencies: The Transformation of Government Guidance. For other posts in the series, click here.

We live in a computerized, point-and-click world. Increasingly, this means that our interactions with the government also are reduced to pointing and clicking at the computer.

In many ways, this is a good thing. Point-and-click access to understanding your legal rights and obligations is not only quick and efficient, it is democratizing. Once upon a time, understanding your legal rights and obligations as a citizen and a taxpayer required hiring professional assistance or, if you couldn’t afford or just didn’t want to pay to do that, visiting a library or a government reading room to read hard copy materials that you might struggle to understand. Now you can point and click to get the guidance you need, and not just in the form of static text versions or PDFs of hard copy materials, but rather by interacting with automated guidance tools seemingly designed to provide quick and easy answers to your questions. With greater access comes greater awareness and understanding, at least in theory.

Except—and this is a big, big caveat—what if the answers we get from point-and-click government are wrong?  In their excellent book, Automated Agencies: The Transformation of Government Guidance, Josh Blank and Leigh Osofsky demonstrate this is no mere academic question, but all too often reflects reality on the ground.  

To some extent, wrong answers come from what Blank and Osofsky label “simplexity,” which is “when the government presents clear and simple explanations of law that is, in fact, ambiguous and complex.” Those of us who routinely interact with complex statutory and regulatory schemes readily appreciate that reducing complicated legal requirements to easy-to-understand instructions for ordinary citizens is really hard.  Tradeoffs between accuracy and readability are undoubtedly inevitable.

As further reported by Blank and Osofsky, however, errors in the government’s automated legal guidance also occur because agency officials “focus[] on usability rather than accuracy, because of a belief that users ha[ve] little ability, inclination, or both, to read complex material [and] automated guidance need[s] to offer answers that [are] as simple as possible.”  In interviews conducted by Blank and Osofsky, agency officials showed “little concern regarding the ways that their guidance may be a poor fit in a given circumstance or for how users may be relying on the guidance in ways that may ultimately leave users vulnerable.” Blank and Osofsky also “heard minimal concern about how the public may view statements by the government as an authoritative and binding form of law, even if agency officials do not think that the statements fill this role.”  (Chapter 5.)  Even worse, Blank and Osofsky found that agencies’ automated legal guidance typically lacked any sort of notice or disclaimer that users should not rely on the answers received (Chapters 7 & 9)—despite repeated recommendations from the Administrative Conference of the United States (ACUS) that agencies should signal when their guidance is not binding or otherwise describe the legal effects of guidance on members of the public.  (See, e.g., ACUS Statement of Principles for Agency Guidance Documents; ACUS Recommendations 2017-5 and 2019-3.)  Particularly at a time when significant swaths of the American public view administrative governance with greater skepticism or even outright distrust, anyone who cares about the credibility of administrative governance should be alarmed by these findings.

Public demand for agency guidance is insatiable.  Knowing how agencies interpret the laws that they administer and would apply those laws to various fact patterns is incredibly useful—whether one is planning transactions, applying for government benefits, or just trying to follow the law on a day-to-day basis.  Prudent regulated parties and potential recipients of government benefits value whatever scraps of information they can get about a relevant agency’s thinking.

The Administrative Procedure Act provides exceptions from notice-and-comment rulemaking procedures for interpretative rules and policy statements precisely so that agencies can be more open in communicating their informal, nonbinding thoughts, plans, understandings, and intentions to the public. Blank and Osofsky question whether agency explanations of law contained in automated guidance fit neatly within the APA’s legislative rule, interpretative rule, or policy statement categories. (Chapter 8.) Regardless, sophisticated regulated parties and the lawyers or other professionals who represent them understand that, however helpful these pronouncements might be, they are nonbinding and, consequently, are only so reliable. Before the internet, when those were the primary users of agency subregulatory guidance, agencies could more readily rely upon that understanding.

Today, that assumption is simply no longer plausible. Agency officials know that ordinary members of the public are reading and relying on their nonbinding pronouncements. Agencies design automated guidance tools specifically for the purpose of serving ordinary members of the public, not sophisticated experts. Agency officials are being willfully blind when they fail to appreciate that members of the public take the pronouncements of automated guidance at face value and rely on them in organizing their affairs.  When the consequences of that reliance prove harsh, public faith in government and the rule of law suffers.

Blank and Osofsky discuss in some depth the implications of their findings for access to justice (Chapter 7) as well as the “democracy deficit” that results from what they describe as a bias toward sophisticated parties in administrative law (Chapter 8).  These consequences are particularly depressing given the tremendous potential of technology for improving public access to and participation in the development of agency guidance. The solutions Blank and Osofsky propose, on the other hand, echo longstanding administrative law goals of increasing transparency and public participation in administrative governance. (Chapters 9 and 10.)  Judicial review is too sparse and too blunt an instrument to solve the problem that Blank and Osofsky identify.  But mechanisms for achieving greater notice and public engagement feature prominently in their suggestions for reform.  Blank and Osofsky also stress the importance for rule of law values of agencies accepting responsibility for public reliance on their pronouncements by “self-bind[ing]” themselves to the pronouncements of their automated legal guidance. (Chapter 10.)  At a minimum, agencies should offer relief from penalties for noncompliance to regulated parties who can demonstrate reliance on their automated legal guidance.

Administering the law has never been easy.  Agencies now are struggling under the weight of hostile political rhetoric and public skepticism that is often unfair.  To a great extent, however, the issues with automated legal guidance that Blank and Osofsky identify reflect agencies shooting themselves in the foot.  Automated Agencies offers smart and concrete proposals for fixing the problem.  We can only hope that agencies will listen.

Kristin Hickman is the McKnight Presidential Professor in Law at the University of Minnesota Law School.