Symposium: Furman’s Legacy: New Challenges to the Overbreadth of Capital Punishment

Download the PDF

Jeffrey Fagan is the Isidor and Seville Sulzbacher Professor of Law at Columbia Law School and a Professor of Epidemiology at the Mailman School of Public Health, Columbia University.


A 2018 decision in the Arizona Supreme Court raised new strong claims that the death penalty in the U.S. has become a “fatal lottery,”[1] with critical implications for its constitutionality and its future in American criminal law.[2] In the case, Hidalgo v. Arizona, the defense provided preliminary evidence that over the past twenty years, nearly 98% of all first- and second-degree murder defendants in Maricopa County—the state’s largest county and location of the nation’s fifth largest city—were death-eligible.[3] The Arizona Supreme Court conceded this point even as it rejected Mr. Hidalgo’s appeal.[4] What the Arizona Supreme Court conceded, and what the evidence showed, was the expansive criteria for death eligibility made it impossible for states to “perform the ‘constitutionally necessary’ narrowing function at the stage of legislative definition” to prevent “a pattern of arbitrary and capricious sentencing.”[5]

Nearly fifty years ago, in Furman v. Georgia, the U.S. Supreme Court cited these same conditions as violating the Eighth Amendment’s cruel and unusual punishment clause to rule the nation’s death penalty statutes unconstitutional.[6] This overbreadth is exactly the opposite of the constitutional requirements set forth over fifty years ago in Furman[7] and four years later in Gregg,[8] seminal U.S. Supreme Court decisions that changed the landscape of capital punishment and created the architecture of the modern death penalty. These cases sought to avoid not only arbitrary but racist outcomes by narrowing capital punishment to a very small subset of cases.[9]

Following the denial by the Arizona Supreme Court but citing its acceptance of the validity of his evidentiary claim, Mr. Hidalgo turned to the U.S. Supreme Court for a review of Arizona’s capital sentencing statute. He again advanced his claim that with so many aggravating circumstances, almost every defendant convicted of first-degree murder would be eligible for the death penalty, a gross violation of Furman’s narrowing requirement and in violation of the Eighth Amendment’s cruel and unusual punishment clause.[10]

Although the Supreme Court declined to take the Hidalgo case,[11] Justice Breyer and three other Justices issued a statement calling for further analysis of whether states have complied with the narrowing requirements set forth in the two core cases of Furman and Gregg.[12] Breyer’s statement noted that “evidence of this kind warrants careful attention and evaluation.”[13] He went on to say that “capital defendants may have the opportunity to fully develop a record with the kind of empirical evidence” that can put these claims to a constitutional test.[14]

Justice Breyer’s Statement signaled that four sitting Justices shared deep concerns about whether “states perform the ‘constitutionally necessary’ narrowing function at the stage of legislative definition” to prevent “a pattern of arbitrary and capricious sentencing.”[15] The Statement went a step further, suggesting a willingness to ask whether statutes, in their operation, are constitutionally suspect, and to apply empirical evidence to address this question. The Furman questions have expanded since the resumption of executions following Gregg. What now matters is not just the statutory architecture of death eligibility—the number of aggravators—but also their scope or reach and their ability to narrow to distinguish “the few cases in which [the death penalty] is imposed from the many cases in which it is not.”[16]

Developing the fact record that Justice Breyer suggests would provide the basis for assessing the constitutionality of a death penalty statute requires a set of thorough, well-designed empirical studies of potentially death-eligible homicides spanning several decades and across several statutory contexts. Determining the breadth of Arizona’s and other states’ statutes requires analyses of the underlying facts of thousands of homicide cases to estimate the rate of death eligibility among them. Because there is no centralized repository of this information, these studies require the collection of records from multiple courthouses and law enforcement agencies, systematic encoding of the information, and analyses tailored to specific statutory eras to determine how broadly the statute operates. It is a daunting challenge, but one that a community of scholars is prepared to meet. The essays in this Symposium are a first step in that direction.

Introduction to the Symposium

This Symposium introduces new research from death sentencing states and local jurisdictions to begin the task of meeting Justice Breyer’s challenge. Researchers and legal scholars convened at Columbia Law School in October 2019 to present empirical and doctrinal scholarship that examines the extent and sources of the overbreadth of capital statutes that was shown fifty years ago in Furman. That overbreadth has re-emerged to show that the failure to narrow is endemic in many of the nation’s death sentencing statutes. These contributions illustrate several features of both statutory design and the institutional practices that replicate the conditions cited by the Furman court to produce regimes of overbreadth, arbitrariness and racial and ethnic disparities.

Professors Catherine Grosso, Barbara O’Brien, and Julie Roberts follow the blueprint designed by Anthony Amsterdam in 2007 in this law review[17] to develop a thick case study of death charging and sentencing practices in Hamilton County, Ohio.[18] The county, which includes Cincinnati, sits on the southern border of Ohio, across the Ohio River from Kentucky. Its history reflects a set of customs and social structures that span both the southern U.S. states and their industrialized northern counterparts. The authors show that over a twenty-five-year period, racial discrimination combines with diffuse statutory eligibility criteria to animate and instantiate the twin concerns of the Furman court: arbitrary death sentences that are imposed in a discriminatory pattern on African-American defendants. Drawing on a historical record and longstanding patterns of discrimination by police and courts, their analysis shows the endogeneity of capital punishment and racial bias in everyday practices in the courts, built on a racially troubled policing regime. Their explanation of the sources and extent of disparate death-seeking follows what Professor Amsterdam envisioned in his call for analyses of death penalty statutes and practices in a rich and deeply contextualized manner.

Hannah Gorman and Margot Ravenscroft, each both a litigator and advocate, remind us that Florida has been among the most aggressive death sentencing states since Furman, and also among the most controversial.[19] Its record of legislative activism created a one-way ratchet to expand death eligibility starting almost immediately after the 1972 Furman ruling. Florida’s statute includes twenty-six enumerated aggravators,[20] and it was one of the first states to create death eligibility for drug delivery in a death.[21] Florida’s patterns of death sentencing, exonerations, and Supreme Court interventions set it apart from nearly every other death sentencing state. Florida’s legislature has battled to retain its unrealistic and rigid view of intellectual disability, and delegates the narrowing function to the prosecutor, not the legislature, in a statutory design similar to the defects that Justice Breyer cited in Hidalgo.[22] Gorman and Ravenscroft reveal empirically how the extent of regional disparity, exploitation by prosecutors of the non-unanimity requirement and statutory expansiveness, and the failure to regulate juror misunderstanding of such basic elements of law as mitigation, create a picture of a de-regulated death penalty system and a dense matrix of Furman problems.

Alexis Hoag is a litigator with deep experience in the convergence of race and arbitrariness in regimes of capital punishment in the U.S. Her doctrinal contribution to the Symposium locates the overbreadth of capital punishment with its seemingly endemic racial disparity in Fourteenth Amendment equal protection doctrine.[23] She departs from the robust claims of racial bias in charging and sentencing of Black defendants, empirical claims that have been muted as constitutional bases of discrimination in the three decades since McCleskey v. Kemp[24] shut down such claims absent a smoking gun of intentional bias.[25] Hoag pivots to the robust empirical evidence of bias in charging and sentencing of killers of White victims, and the inattention by prosecutors in charging defendants of all races and ethnicities to justice for those victims.[26] This devaluation of Black life demands a constitutional remedy under Equal Protection, over and above the Eighth Amendment protections against arbitrary and capricious death sentences. This diminution of the value of life is the essence of Equal Protection law. But the rush to balance these scales without a surgical reduction in eligibility would inevitably worsen the problems of arbitrariness that infect the modern death penalty. For Hoag, abolition of the death penalty is the answer to resolve this tension and balance the values of all lives.

Professor Mona Lynch cites two constitutional flaws in California’s expansive list of “special circumstances,” or statutory aggravators.[27] First is the startling overbreadth of the California death eligibility statute.[28] If the legislature is responsible for the narrowing task proscribed by Furman,[29] then California has failed spectacularly.[30] The breadth of these eligibility factors creates an extraordinarily broad and heterogeneous defendant population. This leads directly to the second problem: the “messier” practice assigned to juries to decide whether the presence of one or more of these circumstances merits a death sentence. Lynch shows how jurors are “swamped” by these multiple indicia of death eligibility, and exerts undue influence on the jury’s sentencing decision. The breadth of these factors burdens jurors who then have to weight these expansive and standardless criteria against mitigation evidence. That jurors have a hard time understanding and applying the law, and making life or death decisions, is revealed in startling and troubling results of a unique set of empirical studies.

Professors Scott Phillips and Trent Steidley expand on Phillips’ earlier work on Texas’ “fatal lottery” to develop evidence of a “systematic lottery” where certain victim-offender killings are systematically declared death eligible, while victim-offender cases are often ignored.[31] They restate the Furman and Hidalgo claims of arbitrariness and overbreadth as two sides of the same coin: death sentencing as so rare as to be “virtually random,” and “yet death sentences are patterned by the race and gender of the victim.” Phillips and Steidley make the trenchant insight into the intersection of the Hidalgo and Furman claims: that the death penalty can be indiscriminate and discriminatory at the same time. The work takes on added importance by focusing on Texas, the most active death sentencing and execution state in the U.S. since reinstatement of capital punishment following Gregg.[32]

A critical implication of the Hidalgo litigation is its reliance on a single-county case study of Maricopa County.[33] Despite the limitations in the evidence record in Hidalgo, Justice Breyer’s statement, signed by three other Justices, suggests that a single-county case study can have constitutional weight in the jurisprudence of capital punishment. Professors Steven Shatz, Michael Pierce, and Glenn Radelet provide evidence—from the largest single-county case study to date—of systematic bias in charging and sentencing, patterns that replicate the statewide evidence in McCleskey over thirty years ago showing particular bias in cases of Black defendants killing White victims.[34] Shatz and colleagues point out that the McCleskey court was amenable to “a sufficiently large single-county study” that can reproduce the statewide findings in that case. The patterns of bias and overbreadth in San Diego County align with Professor Lynch’s showing of the potential for bias and error in the capacious death eligibility criteria in California. These discoveries, when viewed along Professor Grosso and colleagues’ showing in Hamilton County (Ohio), begin to form what Amsterdam envisioned in his original blueprint:[35] a link between social contexts and empirical analyses to show an emerging pattern of county-level constitutional as applied defects in the administration of the death penalty.[36]

Prior to the repeal of Colorado’s death penalty statute[37] in March 2020,[38] research on overbreadth and racial discrimination in charging and sentencing in the state had revealed the pattern of constitutional defects that the Furman[39] Court had warned against nearly 50 years ago.[40] The essay in this volume by Professors Sam Kamin and Justin Marceau, Hidalgo v. Arizona and Non-Narrowing Challenges,[41] reveals not only the presence of the Furman defects of capricious, arbitrary and biased death sentencing in Colorado, but the presence of the same conditions of overbreadth that plagued the Arizona statute discussed by Justice Breyer in his statement in the denial of certiorari in Hidalgo.[42] Kamin and Marceau join the Furman and Hidalgo challenges to provide a blueprint for a state-level challenge highlighting the insurmountable obstacles to resolving the defects cited in Furman and the aspirations of the Gregg[43] design to remedy those flaws. Their focus on the capacity of a death statute to narrow provides the blueprint for the future empirical work, where the fundamental empirical facts about a statute’s inability to narrow are transparent and are blended with the trial facts that draw directly on the defective statute, to provide a record that can only be denied if a court is willing to simply set aside its own constitutional foundations and precedents.

In the final essay in the Symposium, Restoring Empirical Evidence to the Pursuit of Evenhanded Capital Sentencing, Joseph Perkovich, a capital defense attorney, revisits the Hidalgo holding in the Arizona Supreme Court[44] and the denial of certiorari by the U.S. Supreme Court.[45] He links Justice Breyer’s call for an empirical assessment of the constitutional weight of statistical evidence to the precedents set in two earlier U.S. Supreme Court precedents that perhaps should have but didn’t turn on statistical evidence: Lockhart v. McCree[46] and McCleskey v. Kemp.[47] In each case, the Court turned a blind eye to evidence that it otherwise accepted as “methodologically valid”: Lockhart on stacking the deck in jury composition with death-inclined jurors[48] and McCleskey on racial discrimination by prosecutors in their decisions to seek the death penalty.[49] Like the evidence in Hidalgo, the facts in these cases carried enormous weight in the constitutional adjudication of capital punishment, but were swept away with some animus by the Court. Perkovich calls for the reversal of the Court’s anti-science hostility toward the types of robust evidence proffered in each of these cases, asking instead for an open-minded and neutral embrace of the types of complex statistical evidence and experimentation that Justice Breyer and the other justices seek to apply in Hidalgo. His solution goes beyond the Hidalgo episode to create a place for carefully empirically crafted adjudicative facts to bear strong weight in resolving colorable constitutional claims on the death penalty. In doing so, Perkovich returns us to Professor Amsterdam’s blueprint not just on claims of race bias, but on the necessity for a rich and deep body of empirical evidence to resolve constitutional challenges to capital punishment.[50]

We are at a unique and critical moment in the future of the death penalty in the United States. Justice Breyer has opened a new path for researchers and legal scholars to assess the constitutionality of the death penalty, and to create a space for empirical facts that speak directly to the doctrinal issues in constitutional adjudication of capital punishment that have occupied the Court since well before Furman.[51] The articles in this Symposium shed light on this path, showing the critical intersection of Eighth Amendment arbitrariness and Fourteenth Amendment equal protection violations in the emerging jurisprudence of the death penalty, including the centrality of race in both constitutional defects. These studies present a new way to challenge the basic architecture of the modern death penalty, building on and merging the existing doctrines. They illustrate a paradigm for empirical constitutional research on the modern practice of capital punishment, and whether Furman’s constitutional design can cure what may be incurable flaws.

*         We thank the Academy for Justice at the Sandra Day O’Connor College of Law at Arizona State University for its generous support for the Symposium. The Academy for Justice is dedicated to making non-partisan, fact-based academic research available to broad audiences with the goal of creating a criminal justice system in which actual practices reflect best practices.

[1].  Scott Phillips & Alena Simon, Is the Modern Death Penalty a Fatal Lottery? Texas as a Conservative Test, 3 Laws 85, 92 (2014) (describing the pattern of arbitrary and capricious death sentencing as a “fatal lottery”).

[2].  Petition for Certiorari, Hidalgo v. Arizona, No. 17-251, 2017 WL 3531089 at *1 (Aug. 14, 2017).

[3].  Id. Hidalgo’s defense team submitted empirical evidence showing that of the 866 first degree murder cases prosecuted in Maricopa County between 2002 and 2012, 97.8% were capital-eligible. See Cassia Spohn, Aggravating Circumstances in First-Degree Murder Cases, Maricopa County, AZ: 2002–2012 (2018), []. Two different versions of the Arizona statute, one with 10 factors and a second with 14 factors, failed to perform the constitutionally required narrowing.

[4].  State v. Hidalgo, 390 P.3d 783, 791 (Ariz. 2017) (assuming that “Hidalgo is right in his factual assertion that nearly every charged first degree murder could support at least one aggravating circumstance”).

[5].  Id. See also Hidalgo v. Arizona, 138 S. Ct. 1054, 1057 (2018) (Breyer, J., statement respecting the denial of certiorari) (quoting Zant v. Stephens, 462 U.S. 862, 878 (1983)). The Furman Court stated that narrowing was necessary to avoid a pattern of arbitrary and capricious punishments that would violate the Eighth Amendment’s prohibition against cruel and unusual punishment. Furman v. Georgia, 408 U.S. 238, 295 (1972).

[6].  Furman, 408 U.S. at 295; see David C. Baldus, George Woodworth, Michael Laurence, Jeffrey Fagan, Catherine M. Grosso & Richard Newell, Furman at 40: Constitutional Challenges from California’s Failure to (Again) Narrow Death Eligibility, 16 J. Emp. Leg. Stud. 693 (2019).

[7].  Furman, 408 313 (1972) (White, J., concurring) (stating that a death-sentencing procedure is unconstitutional if it provides “no meaningful basis for distinguishing the few cases in which [death] is imposed from the many cases in which it is not.”).

[8].  428 U.S. 153, 189 (1976) (plurality opinion) (“Furman mandates that where discretion is afforded a sentencing body on a matter so grave as the determination of whether a human life should be taken or spared, that discretion must be suitably directed and limited so as to minimize the risk of wholly arbitrary and capricious action.”).

[9].  The Furman Court linked arbitrary patterns of sentencing with racial disparities in sentencing: “It would seem to be incontestable that the death penalty inflicted on one defendant is ‘unusual’ if it discriminates against him by reason of his race, religion, wealth, social position, or class, or if it is imposed under a procedure that gives room for the play of such prejudices.” 408 U.S. at 242 (Douglas, J., concurring); see also Catherine M. Grosso, Jeffrey Fagan, Michael Laurence, David Baldus, George Woodworth & Richard Newell, Death by Stereotype: Race, Ethnicity and California’s Failure to Implement Furman’s Narrowing Requirement, 66 UCLA L. Rev. 1394 (2019) (finding that several of California’s aggravating circumstances are applied disparately based on the race or ethnicity of the defendant).

[10].  Petition for Certiorari, Hidalgo v. Arizona, No. 17-251, 2017 WL 3531089 at *1 (Aug. 14, 2017). Following Gregg, Arizona provided nine statutory aggravators. At the time of Hidalgo’s conviction, the Arizona statute contained 10 aggravators. By the time Mr. Hidalgo filed his petition for certiorari, Arizona had 14. Subsequently, on April 10, 2019, the Arizona Governor signed legislation that removed or significantly modified three of the statutory aggravators: (1) if the defendant created a grave risk of death to another person in addition to the person murdered; (2) if the offense was committed in a cold, calculated manner without pretense of moral or legal justification; and (3) if the defendant used a remote stun gun in the commission of the offense as defined in the statute. See Ariz. Rev. Stat. Ann. § 13-751 (2019).

[11].  Hidalgo v. Arizona, 138 S. Ct. 1054, 1054 (2018).

[12].  Id. (Breyer, J., statement respecting the denial of certiorari).

[13].  Id. at 1057.

[14].  Id.

[15].  Id.

[16].  Furman v. Georgia, 408 U.S. 238, 313 (White, J., concurring); see John Mills, How to Assess the Real World Application of a Capital Sentencing Statute: A Response to Professor Flanders’ Comment, 51 U.C. Davis L. Rev. Online 77, 80 (2017).

[17].  Anthony G. Amsterdam, Opening Remarks: Race and the Death Penalty Before and After McCleskey, 39 Colum. Hum. Rts. L. Rev. 34, 49 (2007).

[18].  Catherine M. Grosso, Barbara O’Brien & Julie C. Roberts, Local History, Practice, and Statistics: A Study on the Influence of Race on the Administration of Capital Punishment in Hamilton County, Ohio (January 1992–August 2017), 51 Colum. Hum. Rts. L. Rev. 904 (2020).

[19].  Hannah L. Gorman & Margot Ravenscroft, Hurricane Florida: The Hot and Cold Fronts of America’s Most Active Death Row, 51 Colum. Hum. Rts. L. Rev. 937 (2020).

[20].  Fla. Stat. § 782.04(1)(a) (2019).

[21].  Fla. Stat. §§ 782.04(3), 775.082 (2019).

[22].  Hidalgo v. Arizona, 138 S. Ct. 1054, 1057 (2018) (Breyer, J., statement respecting the denial of certiorari).

[23].  Alexis Hoag, Valuing Black Lives: A Case for Ending the Death Penalty, 51 Colum. Hum. Rts. L. Rev. 985 (2020).

[24].  McCleskey v. Kemp, 481 U.S. 279 (1987).

[25].  See Amsterdam, supra note 17, at 45–47. See, e.g. Randal Kennedy, McCleskey v. Kemp: Race, Capital Punishment, and the Supreme Court, 101 Harv. L. Rev. 1388, 1392 (1988) (discussing in-group bias in black homicides).

[26].  Hoag is careful to locate the devaluation of Black victim lives but never loses sight of the deep and persistent bias toward Black defendants.

[27].  Mona Lynch, Double Duty: The Amplified Role of Special Circumstances in California’s Capital Punishment System, 51 Colum. Hum. Rts. L. Rev. 1010 (2020).

[28].  David C. Baldus et al., Furman at 40, supra note 6 (showing that over 90% of first degree murder convictions are death eligible under California’s “special circumstances.”).

[29].  Hidalgo, 138 S. Ct. at 1054 (Breyer, J., statement respecting the denial of certiorari) (“To satisfy the ‘narrowing requirement,’ a state legislature must adopt ‘statutory factors which determine death eligibility’ and thereby ‘limit the class of murderers to which the death penalty may be applied.’”).

[30].  See, e.g., Jonathan Simon & Christina Spaulding, Tokens of Our Esteem: Aggravating Factors in the Era of Deregulated Death Penalties, in The Killing State: Capital Punishment in Law, Politics, and Culture 81, 81 (Austin Sarat ed., 1999) (describing the continuous expansion of death eligibility by the California legislature for over a decade beginning with the reinstatement of the death penalty in 1977).

[31].  Scott Phillips & Trent Steidley, A Systematic Lottery: The Texas Death Penalty, 1976 to 2016, 51 Colum. Hum. Rts. L. Rev. 1043 (2020).

[32].  Gregg v. Georgia, 428 U.S. 153, 189 (1976) (plurality opinion); see Death Sentences in the United States Since 1977, Death Penalty Info. Ctr., []; Executions Overview, Death Penalty Info. Ctr., https://deathpenalty [].

[33].  The city of Phoenix in Maricopa County is the fifth most populous city in the U.S., the largest state capital, and the only state capital with a population of more than one million residents. Its land area is greater than New York, Los Angeles, or Chicago. Bernard Goth, Take That, Philly: Phoenix Reclaims the Title of 5th-Largest U.S. City, Republic (May 25, 2017), [].

[34].  Steven F. Shatz, Glenn L. Pierce & Michael L. Radelet, Race, Ethnicity, and the Death Penalty in San Diego County: The Predictable Consequences of Excessive Discretion, 51 Colum. Hum. Rts. L. Rev. 1072 (2020).

[35].  Amsterdam, supra note 17.

[36].  See Baldus et al., supra note 6; Grosso et al., supra note 9.

[37].  Colorado SB20-100 repealed the death penalty for all previously death-eligible crimes committed on or after July 1, 2020.

[38].  Andrew Kenney, Colorado Death Penalty Abolished, Polis Commutes Sentences of Death Row Inmates, Colorado Public Radio (Mar. 23, 2020), [].

[39].  Furman v. Georgia, 408 U.S. 238, 295 (1972). See Baldus et al., Furman at 40, supra note 6.

[40].  See, e.g., Meg Beardsley, Sam Kamin, Justin Marceau, & Scott Phillips, Disquieting Discretion: Race, Geography & the Colorado Death Penalty in the First Decade of the Twenty-First Century, 92 Denv. L. Rev. 431 (2015). Prior to repeal, Colorado juries had not handed down any death sentences in over a decade, and the state’s last execution was in 1997. Colorado juries had not imposed any death sentences in a decade, and the state’s last execution was more than 20 years ago, in 1997. In 2013, then-Governor John Hickenlooper imposed a moratorium on executions, calling the state’s death penalty system flawed and inequitable. Colorado Becomes the 22nd State to Abolish the Death Penalty, Death Penalty Info. Ctr. (Mar. 24, 2020), [].

[41].  Sam Kamin & Justin Marceau, Hidalgo v. Arizona and Non-Narrowing Challenges, 51 Colum. Hum. Rts. L. Rev. 1101 (2020).

[42].  Hidalgo v. Arizona, 138 S. Ct. 1054, 1054 (2018) (statement of Breyer, J., respecting the denial of certiorari).

[43].  Gregg v Georgia, 428 U.S. 153, 189 (1976) (plurality opinion).

[44].  State v. Hidalgo, 390 P.3d 783 (Ariz. 2017).

[45].  Hidalgo, 138 S. Ct. at 1054.

[46].  476 U.S. 162 (1986).

[47].  481 U.S. 270 (1987).

[48].  476 U.S. at 173.

[49].  McCleskey v. Kemp, 481 U.S. 279, 308 (1987).

[50].  Amsterdam, Opening Remarks: Race and the Death Penalty Before and After McCleskey, supra note 17.

[51].  Frederick Schauer & Barbara A. Spellman, Probabilistic Causation in the Law, 176 J. Institutional & Theoretical Economics 4, 13 (2020) (recognizing and arguing for a larger role of empirical evidence and counterfactual reasoning in resolving legal questions).