Data & the Death Penalty: Exploring the Question of National Consensus Against Executing Emerging Adults in Conversation with Andrew Michaels’s A Decent Proposal: Exempting Eighteen‑ to Twenty‑Year‑Olds From the Death Penalty

Brian Eschels


I. National Consensus Through Practice

II. Where Do Courts Look For National Sentencing Statistics?

III. Possible Sources of Offender Age Data

IV. Clark Prosecutor Data

V. Trends in the Execution of Emerging Adults

In A Decent Proposal: Exempting Eighteen‑ to Twenty‑Year‑Olds From the Death Penalty,[1] Andrew Michaels argues that emerging adults—those between the ages of eighteen and twenty—should be categorically exempted from the death penalty. Relying on the Court’s recent line of Eighth Amendment decisions in Atkins v. Virginia,[2] Roper v. Simmons,[3] and Graham v. Florida,[4] Michaels argues that the test for a categorical finding of cruel and unusual punishment has been satisfied: (1) there is a national consensus against the execution of emerging adults, and (2) the application of the death penalty to emerging adults is disproportionate to their culpability. Michaels argues that society recognizes the relative immaturity—and thus, reduced culpability—of emerging adults by erecting legal regimes that treat emerging adults differently than older adults, particularly in the areas of firearms, alcohol, and foster parenting. In addition, psychological and neurological studies have long recognized that the brain is not fully developed until the mid‑twenties, meaning that emerging adults are more prone to risk‑seeking behavior and less able to resist peer pressure. Michaels’s article adeptly and convincingly presents this argument. The purpose of this article is concerned exclusively with addressing the first prong of the test: is there a national consensus against the execution of emerging adults?

Testing this claim necessarily involves presenting not only statistical conclusions, but also the underlying data from which those conclusions are drawn. The print format does not easily allow authors to append spreadsheets of data, and even if it did, readers would be unable to inspect the spreadsheet’s internal functions. That is why this article, rather than A Decent Proposal, presents the statistical evidence for a national consensus against the execution of emerging adults. Specifically, this article examines what kinds of evidence courts look to in order to find a national consensus against a certain punishment, the available sources of data cataloging the execution of emerging adults, and what conclusions can be drawn from that data. Appendix I presents the data itself in an easy‑to‑use excel sheet.

I. National Consensus Through Practice

In finding a national consensus against a certain punishment, the Court has weighed the number of states in which that punishment is lawful, as well as the frequency of the punishment’s application.[5] In Graham v. Florida, the Court addressed whether the Eighth Amendment prohibits a sentence of life without parole (LWOP) for juvenile non‑homicide offenders.[6] The Court answered in the affirmative and, in doing so, found that infrequency of application alone is enough to find a national consensus against a punishment, even if that punishment is a legally available sentencing option in a majority of states.[7] In Graham, the Court noted that while forty‑four states allowed LWOP sentences for juvenile non‑homicide offenders, “an examination of actual sentencing practices in jurisdictions where the sentence in question is permitted by statute discloses a consensus against its use.”[8] At that time, only 123 juvenile non‑homicide convicts were imprisoned under LWOP sentences, spread among just eleven jurisdictions.[9] The Court concluded that these numbers were low enough to establish a national consensus against life sentences without parole for non‑homicide juvenile offenders.[10]

As was true in Graham, a large number of states permit the execution of emerging adults by statute.[11] Thus, any consensus against such executions will have to be found in the frequency of the practice when viewed “in proportion to the opportunities for its imposition.”[12]

II. Where Do Courts Look for National Sentencing Statistics?

When the Court in Roper v. Simmons categorically struck down the juvenile death penalty, it noted that in the ten years prior, only three states had executed a juvenile.[13] The court cited to the 76th volume of a periodic report on the juvenile death penalty put forth by Professor Victor Streib of the Pettit College of Law at Ohio Northern University.[14] That report is very thorough, but did not cite any sources in support of its factual assertions. Instead, it contained an appendix listing the name, age at offense, race, state, and disposition for every juvenile sentence of death imposed from 1973 through 2004,[15] as well as an appendix presenting a short descriptive paragraph for each person under a sentence of death as of April 30, 2004.[16] The court’s faith in the accuracy of the report’s assertions appears to be based solely on Professor Streib’s reputation as an interested scholar‑advocate and the fact that he released seventy‑seven such reports on the topic of the juvenile death penalty alone.[17] In total, Professor Streib’s reports have been cited by the Supreme Court twenty‑eight times according to his biography,[18] including twelve times in Thompson v. Oklahoma, where he was on the brief.[19]

A second prominent source of statistics for the Court are those provided by the Death Penalty Information Center (DPIC), which has also been cited numerous times by federal and state courts.[20] Occasionally, the Court has relied on single, non‑recurring studies that are not affiliated with a state or nonprofit. In Graham, the leading case for finding national consensus through practice, the court relied on such a study cataloguing the number of juvenile non‑homicide offenders serving life sentences.[21] That study collected data through public records requests,[22] but did not present the raw data. In addition, the report’s data was incomplete at the time of Graham (due to late FOIA responses) and was not peer reviewed.[23] In sum, the Court seems to be grasping for whatever statistics are available, and is willing to overlook certain limitations, such as lack of source data or peer review, when a certain source constitutes the best option available.

III. Possible Sources of Offender Age Data

Tracking down the age of each executed person at the time of the offense is more difficult than it sounds. Professor Streib’s reports never collected this information for all offenders, and he has since retired.[24] The Bureau of Justice Statistics (BJS) does not track this number, and neither does DPIC’s main database, the go‑to source of statistics for the capital defense bar. [25]

The three sole sources of aggregated national execution data that include age at the time of offense appear to be (1), a website maintained by journalist Juan Ignacio Blanco, (2) a defunct excel spreadsheet created by Professor Dean Golding, which is available through a search function on DPIC’s website,[26] and (3) the website of the prosecutor for Clark County, Indiana (“Clark Prosecutor”).[27]

The first two sources explicitly rely on the third, listing Clark Prosecutor as a source of their data. Each of the sources is incomplete in a unique way. The Golding spreadsheet only lists data for executions preceding Joseph Martin Luther Gardner’s execution on December 5, 2008.[28] does not list any executions after Lisa Ann Coleman was executed on September 17, 2014.[29] The Clark Prosecutor data is only slightly more complete, listing executions up to Leon Vincent Taylor’s execution on November 19, 2014.[30] In addition, none of the datasets list date of offense or age at the time of offense for persons executed before 2000.[31]

Given that Clark Prosecutor appears to be the original source of date‑of‑offense data, I will move forward using that dataset as my starting point.

IV. Clark Prosecutor Data

There are several indicia of reliability for the Clark Prosecutor data. First, the site maintains a factual summary page for each executed person and cites facts from specific states’ Corrections Department pages, as well as other sources.[32] Second, a comparison between DPIC’s list of executed persons between 1977 and November 19, 2014, and Clark Prosecutor’s shows 100% agreement as to the number of persons executed, as well as their names.[33] Third, DPIC—an organization that the Supreme Court has relied on—used Clark Prosecutor as a source for the previously mentioned spreadsheet listing certain data about persons executed between 1977 and 2008.[34]

The Clark Prosecutor data do not list any executions after November 19, 2014. With the goal of completing the dataset through the end of 2015, this author imported the names of all persons executed between November 19, 2014 and before January 1, 2016 from DPIC’s database, and found corroborating sources for these person’s dates of birth, states of execution, and dates of offense. There were thirty executions in this time period and these were added to the existing Clark Prosecutor data. While Clark Prosecutor data are messy to begin with (as they are scraped from a website, rather than downloaded in an easy to use format), this author managed to organize them in a useful way and extract pertinent aggregate statistics.

V. Trends in the Execution of Emerging Adults

After some careful analysis, the data reveal that executions of emerging adults are rare and occur in just a few states, especially in comparison to the prolific homicide rate among emerging adults. In 2010, emerging adults lead in BJS data cataloguing murders and non‑negligent homicides, with nineteen‑year‑olds killing 744 and eighteen‑year‑olds killing 709.[35] In contrast, twenty‑four year olds killed 449.[36] If there were no national consensus against executing emerging adults, one would expect that the practice of executing members of this high‑violence group would be common. It is not.

From 2001 to 2015, twenty‑eight states executed at least one adult, but only fifteen states executed anyone between eighteen and twenty.[37] Over those fifteen years, a total of 130 emerging adults were executed, in comparison to 730 people (excluding pre‑Roper juveniles) executed in total. Of those emerging adults, 77.69% of them were executed in just four states—Texas, Oklahoma, Virginia, and Ohio—whereas the top four full‑adult executing states (which were the same, save the substitution of Florida for Virginia) accounted for only 61.17% of executions.[38] Texas alone accounted for 55.38% of all emerging adult executions, but only 36% of full‑adult executions.[39]

In other words, the practice of executing emerging adults appears to be localized in just a handful of states and that localization is not entirely explained by general trends in execution frequency. In the last five years (2011–2015), the concentration of emerging adult executions has grown, even as the overall practice has become less frequent. Only nine states executed an emerging adult in that time, with Texas executing eighteen (representing 58.06% of all executions of this age group) emerging adults over those years.[40] The next highest state, Georgia, executed just three emerging adults.[41] In those five years, only thirty‑one emerging adults were executed, out of 187 total executions.[42] In 2015, only Texas executed any emerging adults at all, killing five.[43]

In Graham, the Court found a national consensus against LWOP for non‑homicide juvenile offenders based on the fact that only eleven states had any juveniles serving such sentences, and 77 of 123 (62.6%) of those sentences were imposed in Florida.[44] The Clark Prosecutor data for the last five years closely mimic this distribution, with only nine states participating in the practice and one state, Texas, accounting for 58.06% of all activity.[45] Of course, the comparison is slightly inexact, as the LWOP sentences noted in Graham may have been handed down over a period longer than five years. Regardless, the pattern of state participation in this punishment has come to mimic the pattern the court used in Graham to strike down juvenile LWOP for non‑homicide offenses.

Andrew Michaels argues in A Decent Proposal that this national consensus, combined with the disproportionality of the death penalty in relation to the culpability of emerging adults, compels the conclusion that the death penalty is categorically cruel and unusual as applied to this group. The data reported here support the idea that there is a national consensus against the execution of emerging adults.



1. 40 N.Y.U. Rev. L. & Soc. Change ___ (2016).

2. Atkins v. Virginia, 536 U.S. 304 (2002).

3. Roper v. Simmons, 543 U.S. 551 (2005).

4. Graham v. Florida, 560 U.S. 48 (2010).

5. Id. at 62; see also Enmund v. Florida, 458 U.S. 782, 794 (1982) (“Society’s rejection of the death penalty for accomplice liability in felony murders is also indicated by the sentencing decisions that juries have made”).

6. Graham, 560 U.S. at 52–53.

7. Id. at 82, 62–67. The Court has since gone on to hold that mandatory LWOP sentences for juvenile homicide offenders likewise constitutes cruel and unusual punishment. Miller v. Alabama, 132 S. Ct. 2455 (2012).

8. Id. at 62.

9. Id. at 64.

10. Id. at 67.

11. Specifically, the execution of emerging adults is legal in all states with active death penalty statutes. Nineteen states have effectively ended capital punishment within their borders. See Michaels, supra, note 1, at ____ n.178 (reporting that sixteen states have banned the death penalty in its entirety, two more states have banned the imposition of new death penalties without resolving the disposition of those currently on death row, and one state has effectively ended the death penalty by declining to replace its capital punishment statute which whose application (both prospective and retrospective) was found unconstitutional)

12. See Graham, 560 U.S. at 66.

13. Roper, 543 U.S. at 565 .

14. Id. See Victor Streib, The Juvenile Death Penalty Today: Death Sentences and Executions for Juvenile Crimes, January 1, 1973–December 31, 2004, Issue No. 76 (2004).

15. Streib, The Juvenile Death Penalty Today, supra note 15 at app. A.

16. Id. at app. B.

17. Professor Streib was on the brief for petitioner in Thompson v. Oklahoma, 487 U.S. 815 (1988), as well as two state supreme court cases striking down the death penalty for fifteen‑year old offenders. Allen v. State, 636 So.2d 494 (Fla. 1994); Cooper v. State, 540 N.E.2d 1216 (Ind. 1989). Additionally, Professor Streib has written prolifically on the death penalty in general. See, e.g., Victor Streib, Death Penalty in a Nutshell (4th ed. 2012); Victor Streib, Death Penalty for Female Offenders, January 1, 1973 Through October 31, 2010, Issue. No 65, See generally, Ohio Northern University Law Review, Tribute, A Tribute to Victor Streib, 38 Ohio N.U. L. Rev. 405 (2012) (dedicating issue to Professor Streib).

18. Victor Streib, Faculty Profile, ONU Law (last visited Apr. 6, 2016),

19. See Thompson, 487 U.S. 815.

20. See, e.g., Atkins, 536 U.S. 304, 316 n.20, 346, (2002) (Scalia, J., dissenting) (citing a report “updated by Death Penalty Information Center); Kennedy v. Louisiana, 554 U.S. 407, 460 n.7 (2008) (Alito, J., dissenting) (citing DPIC report titled “Executions in the U.S. 1608–2002: The ESPY File Executions by Date (2007)”), Workman v. Bredesen, 486 F.3d 896, 907 (6th Cir. 2007) (citing DPIC webpage collecting facts about lethal injection), State v. Santiago, 318 Conn. 1, 190 (2015) (Eveleigh, J. concurring); Gibson v. Turpin, 270 Ga. 855, 872 n.38 (1999) (Fletcher, P.J., dissenting) (citing DPIC website listing number of death row inmates by state).

21. Paolo G. Annino, David W. Rasmussen & Chelsea Boehme Rice, Juvenile Life Without Parole for Non‑Homicide Offenses: Florida Compared to Nation (updated Sept. 14, 2009),‑content/uploads/2010/06/FSU‑nonhomicide‑report.pdf.

22. See id. at App. II.

23. Graham, 560 U.S. at 62–63 (citation omitted).

24. Scott D. Gerber, Introduction to the Tribute to Victor L. Streib: An Ideal Role Model, 38 Ohio N.U. L. Rev. 407 (2012) (noting Professor Streib’s retirement).

25. Bureau of Justice Statistics, (last visited Feb. 8, 2016); Death Penalty Information Center, (last visited Feb. 8, 2016).

26. Executions in the United States, Death Penalty Information Center,‑united‑states (last visited Apr. 16, 2016) [hereinafter “DPIC Spreadsheet”] (offering the spreadsheet as a resource, but cautioning that “[a]s these statistics were created by a source outside of DPIC, DPIC cannot vouch for its accuracy”).

27. Death Penalty USA, (last visited Feb. 8, 2016); The Death Penalty, Clark Prosecutor, (last visited Feb. 8, 2016), This position is currently held by Jeremy Mull, who took office on January 1, 2015. Jeremy Mull Bio, Clark Prosecutor (last visited Feb. 8, 2016),

28. DPIC spreadsheet, supra note 26.

29. Executions by Date: 2014, DeathPenaltyUSA (last visited Feb. 8, 2016)

30. U.S. Executions Since 1976, Clark Prosecutor (last visited Feb. 8, 2016),

31. The DPIC spreadsheet has data for only some people executed before 2000. DPIC spreadsheet.

32. See, e.g., Lisa Ann Coleman, Clark Prosecutor (last visited Feb. 8, 2016), (hyperlinking to numerous sources, including the Texas Department of Criminal Justice’s page listing Lisa Ann Coleman’s demographic information).

33. Comparing two lists of 1,392 names is difficult, but not impossible. This author imported both lists into an excel sheet and used functions to normalize the lists (as Clark Prosecutor was much more likely to list middle names, initials, suffixes, and nicknames than DPIC was) and then automatically compare them. This resulted in the relatively rapid verification of roughly 1,200 names. The remaining names had to be hand‑checked. These names were not matched by function primarily due to spelling disagreement or middle‑name‑as‑alias disagreement. A hand‑check revealed that the names match once the disagreement was resolved.

34. See Faculty Profile, supra note 20; DPIC spreadsheet., supra note 26.

35. Howard N. Snyder, Bureau of Justice Statistics, U.S. Dep’t of Justice, Arrest in the United States, 1990–2010, at 17–18 (2012),

36. Id. at 18.

37. Appendix I.

38. Id.

39. Id.

40. Id.

41. Id.

42. Id.

43. Id.

44. Graham, 560 U.S. at 49.

45. Appendix I.

%d bloggers like this: