The point of the Fourth Amendment, which often is not grasped by zealous officers, is not that it denies law enforcement the support of the usual inferences which reasonable men draw from evidence. Its protection consists in requiring that those inferences be drawn by a neutral and detached magistrate instead of being judged by the officer engaged in the often competitive enterprise of ferreting out crime.
— Justice Robert H. Jackson1 (1948)
Introduction
On March 13, 2020, after breaking her door down with a battering ram, seven officers flooded Breonna Taylor’s Louisville home and shot her five times.2 Two years later, on February 2, 2022, SWAT officers from the Minneapolis Police Department entered an apartment where Amir Locke was staying, and when he emerged from under a blanket with a gun, an officer shot him three times, killing the 22-year-old less than ten seconds after law enforcement set foot in the home.3 Locke was not named in the search warrant authorizing the raid of the location where he was killed, nor were either Locke or Taylor suspects in the police investigations motivating the searches that led to their deaths.4
Although stories of police shootings seem so common as to almost become unremarkable,5 the Taylor and Locke killings provoked a particularly strong outcry from the public due in large part to the fact that neither death was merely the result of spur-of-the-moment decisions by law enforcement.6 On the contrary, the factual background and legal basis for each of these fatal entries were preceded by a formal, judicial review that could have — and arguably should have — identified the various errors that served as the basis for the initial warrants. Indeed, subsequent investigations into both incidents showed that the submitting officers perjured themselves and the reviewing judges (or magistrates) approved the no-knock warrants when they were questionable at best7 — the precise kinds of behavior that constitutionally prescribed judicial review of warrants is designed to curb.8
The stories of Breonna Taylor and Amir Locke uncover, in the direst way, the repercussions of inaccurate or unconstitutional search warrants. These stories also urge review and reconsideration of search warrants more broadly. Between 2010 and 2016, at least eighty-one civilians and thirteen officers died in “dynamic entry” raids including no-knock raids.9 More commonly, serious injuries and property damage result from searches, requiring millions of dollars of litigation costs and payouts from police departments.10 Even if no death, injury, or property damage results from a warrant, nearly all warrant-based searches impose substantial privacy costs, as officers enter homes, search and detain individuals, and examine digital items like cell phones and computers that contain troves of particularly personal information. And ignoring the individualized costs, the invasiveness of searches inevitably undermines trust in law enforcement, especially in overpoliced, marginalized communities.11
Recognizing the immense scale of police searches and seizures in the United States — with hundreds of thousands, and possibly millions, of searches being performed on individuals and their property12 — the conventional wisdom in Fourth Amendment jurisprudence and much of the academic literature is that the warrant requirement serves as a necessary and legitimate check on police power.13 But a warrant functions as a check in large part (if not exclusively) because the law enforcement officer must submit an affidavit under oath to be vetted by a “neutral and detached magistrate” who can narrow the scope or deny it.14
The doctrine surrounding search and seizure has been subject to significant examination by judges and scholars,15 but the aspects of how warrant review actually works on the ground have received remarkably little attention, leaving critical questions unanswered. Consistent with the constitutional requirements in the Fourth Amendment, we know that neutral judges are required in warrant review, but what role do these judges actually play in such a determination — that is, how are such decisions made in practice, and to what extent do judges actually function as a “neutral and detached” check against “officer[s] engaged in the often competitive enterprise of ferreting out crime?”16 Are the failures demonstrated in the Breonna Taylor and Amir Locke events tragic but unavoidably random anecdotes, or are they the product of systematic institutional failures? And how do the answers to these questions color our understanding of the extent to which a warrant requirement truly matters at all? Finally, do these searches actually work in practice?
The public’s underlying concern with the warrant process has borne an influx of critical reforms designed to increase accountability, particularly in the wake of the deaths of Breonna Taylor and Amir Locke.17 Six states — Connecticut, Florida, Oregon, Tennessee, Washington, and Virginia — banned no-knock warrants in their entirety,18 and as of April 2023, thirty states and twenty-five cities have enacted some form of restrictions on no-knock raids.19 Utah (the venue of this Article’s empirical study) requires police officers to identify themselves audibly when using no-knock warrants and bans their use entirely for misdemeanor crimes as of 2022.20 Some three years earlier, in March 2019,21 Utah also enacted the Electronic Information or Data Privacy Act22 (H.B. 57), which required law enforcement to obtain a warrant when requesting information from electronic mail providers, wireless phone companies, search engine firms, and social media companies.23
But while the national appetite for warrant reform has arguably never been higher, actual data on warrants is incredibly scarce. To the best of our knowledge, there have been only four empirical studies on the warrant application and review process conducted in the last forty-five years.24 Three of these were done prior to the arrival of smartphones and the broad proliferation of electronic warrants, and all of them relied on relatively limited datasets. In his 2012 study, Professor Thomas Miles investigated the propensity for “judge shopping”25 in all wiretap warrant submissions to federal judges from 1997–2007, looking specifically to see whether judges with certain ideological leanings, professional backgrounds, or racial/ethnic identities were more likely to receive warrant applications.26 The seminal 1985 study by Richard Van Duizend and others observed eighty-four warrant proceedings and found that 65% of the warrant applications were decided in around two-and-a-half minutes or less, with a median time of two minutes and twelve seconds.27 The reports stemming from the San Diego Search Warrant Study were similarly discouraging, finding probable cause violations and racial disparities among nearly 250 drug-related search warrants issued in 1998.28 More recently, the Salt Lake Tribune evaluated a smaller and more limited version of the dataset we utilized in this Article (their exposé was the inspiration for our study), analyzing 844 warrants approved in Utah over the course of one month in 2017.29 The report accused the judiciary of being “too hasty” with warrant review and sometimes even “rubber stamping” approvals when the content of the affidavit did not meet probable cause standards.30
Most of these studies raised concerns about the speed at which warrants were reviewed and the high rates at which they were approved, questioning whether “the suppositions regarding the Fourth Amendment warrant requirement may not be borne out in practice.”31 However, relying on the conclusions drawn from such small samples (and old data) may paint a picture that is incomplete and potentially misleading. Context is important, and some may argue that the inherently hurried environment in which warrants are sought, combined with the low standard of review, creates a system in which swift reviews and approvals are expected, necessary, and, ultimately, adequate. Similarly, warrant affidavits are short and only require the officer to establish a reasonable basis for believing that evidence for a crime exists,32 so the necessary time for review might be less than critics believe. Furthermore, judges, especially state trial-level judges, often are underresourced, face a growing backlog of potentially more critical matters, and may be incentivized to maintain a positive relationship with law enforcement.33
This Article seeks to fill this empirical gap (at least in part) by analyzing previously untapped statewide data from Utah’s “e-Warrants” system,34 constituting the largest and most comprehensive quantitative study of warrants of any kind. The dataset includes the affidavits of every electronic warrant submission during a nearly three-year period (March 2017 through January 2020),35 totaling 33,465 warrant applications filed by more than 3,200 law enforcement officers and reviewed by nearly 120 judges and magistrates. Even more, the dataset includes metadata captured by the e-Warrants system that chronicles the timing of every important stage of the warrant review process, including when the officer submits the warrant affidavit; when the reviewing judge first opens the submission; the amount of time the judge spends reading the warrant affidavit; the moment the judge makes their legal determination; and if and when the warrant is served. By combining the warrant affidavit texts with digital timestamps and situating that data in the legal and extralegal context of judicial warrant review, we provide the most expansive and comprehensive view into the search and seizure process.
Our findings suggest that the on-the-ground reality of warrant review is often not meeting the robust assumptions that are baked into Fourth Amendment jurisprudence and scholarship. At the same time, the results are also not fully consistent with the skeptical position that judges are not actually reviewing the warrants at all. On the one hand, some (albeit very few) warrants go through multiple revisions before being approved, and we identify moderate correlations between a judge’s professional background and review outcomes — observations that run counter to the proposition that all warrants are simply “rubber stamp[ed].”36 On the other hand, our key findings demonstrate that over 93% are approved on the first attempt; over 98% of all warrants are eventually approved; and the median time for a warrant affidavit to be opened, read, analyzed, and (in the case of approval) sent back to the officer in the form of a judicial warrant is only three minutes. Furthermore, one out of every ten warrants is read, analyzed, and responded to in a minute or less.
Accounting for warrant complexity and length, we determine that at least half of all approved warrants are either not read in full or are skimmed by the reviewing judges. These results are consistent even when accounting for the most common explanations provided to us during interviews with judges who review warrants (including Utah judges), including legal domains that arguably require less stringent analysis (for example, DUI warrants), identical intra-warrant content, the existence of “unnecessary” language, and warrant resubmissions.
Looking at judge-level characteristics, we also provide some preliminary explorations into the possible factors affecting review time and approval rates. We identify substantial variations across individual judges, even when controlling for district-level factors. In some cases, judges within the same district are fifteen or twenty percentage points more likely to approve warrants and take seven times longer to review them than their colleagues. Regression analyses that control for warrant-, judge-, and time-level factors demonstrate strong (and in some cases surprising) correlations between review time and approval likelihood with a judge’s tenure and previous professional background.
Finally, we also performed a qualitative analysis of a randomly selected subsample of the affidavits and found several troubling examples of questionable judicial review. These include cases where approved warrants were unconstitutionally vague, lacked corroboration of evidence used as the basis for probable cause, or failed to catch important errors in the affidavit.
Taken together, this empirical and qualitative evidence has novel, critical implications for both the theory and praxis of warrant review and execution.
This Article proceeds as follows: Part I provides a brief background on the current state of warrant jurisprudence and practice, laying out an informal behavioral mapping of the extralegal factors that likely influence judicial warrant review. Part II discusses the institutional setting for our empirical study and the nature of the datasets we employ. Parts III and IV present our central empirical analyses, focusing on the time it takes Utah judges to review warrants, the resulting approval rates, and the warrant- and judge-level characteristics that best predict those outcomes. Part V describes a small but in-depth review of individual warrants that provides qualitative evidence of inadequate judicial review. Part VI concludes by discussing the implications of our findings and exploring potential legal, policy, and institutional reforms that address the problems identified in our empirical analyses.
Continue Reading in the Full PDF
∗ Professor and Terry J. Tondro Research Scholar at the University of Connecticut School of Law.
∗∗ Assistant Professor in the Department of Linguistics at Brigham Young University.
∗∗∗ Associate Professor of Law at Brigham Young University Law School. We thank Sam
Aguiar, Jill Anderson, David Armond, Benjamin Berger, Kiel Brennan-Marquez, Samuel Buell,
John de Figueiredo, Timothy Everett, Jeffrey Fagan, Todd Fernow, Eric Fish, Michael Frakes,
Brandon Garrett, Paul Heaton, David Hoffman, Benjamin Johnson, Orin Kerr, Daniel Klerman,
Tom Lee, Kay Levine, Brendan Maher, Jeremy McClane, John Meixner, Pamela Metzger, Murat
Mungan, Palma Paciocco, Lisa Perkins, Jonathan Petkun, J.J. Prescott, Benjamin Pyle, John Rappaport, Richard Re, Ryan Sakoda, David Schwartz, Peter Siegelman, David Sklansky, Noella
Sudbury, François Tanguay-Renaud, Robert Weisberg, Ronald Wright, Kathryn Zeiler, and participants at the American Law & Economics Association (ALEA) 2024 Annual Meeting, Boston University’s Law & Economics Workshop, Duke Law School’s Law & Social Science Workshop, Southern Methodist University’s Deason Criminal Justice Reform Center’s Workshop, Osgoode Hall Law School’s Nathanson Centre Workshop, and the Northwestern Law School/Georgia State University SCALES Workshop for helpful conversations and comments. Zachary Bellis, Judy Chicoine, Brady Earley, Natasha Esponda, Mark Harris, Rhett Hunt, Justin Hyland, Brittany LaMarr, Kyra Nelson, Elisse Newey, Santanna Rocha, Celina Stoia, Curtis Thacker, Teresa White, and John Yun provided incredible programming support and research assistance. We are grateful to Jessica Schreifels and Aubrey Wieber who generously shared their data from investigative work they did for the Salt Lake Tribune. We also thank Brigham Young University Law School, University of Connecticut School of Law, and the SCALES Open Knowledge Network Project at Northwestern Law School for financial support. Finally, and especially, we recognize and thank the Utah judges, law enforcement officers, and court and state agency administrators who helped facilitate this study through providing access to data, doing interviews, and giving feedback, even while recognizing that the results of our analysis might not be flattering.