Use of the Death Penalty for Killing a Child Victim

About half of all death penal­ty states include the mur­der of a child as an aggra­vat­ing cir­cum­stance that can sub­ject a defen­dant to the death penal­ty. As of January 2022, four­teen states autho­rized the death penal­ty for the mur­der of a child vic­tim, and five states that lat­er abol­ished the death penal­ty also had a child-vic­tim aggravating circumstance.

The most com­mon of the age-of-vic­tim require­ments by far — used by nine cur­rent or for­mer death penal­ty states — is that the vic­tim must be under age 12.¹ The next most com­mon age, used by four cur­rent or for­mer death penal­ty states, is that the vic­tim be under age 14.² Two states require that the vic­tim be younger than age 13, and one requires that the vic­tim be younger than age 10.³ Only three states autho­rize cap­i­tal pros­e­cu­tion based upon the vic­tim being 14 years old or old­er.⁴

¹ Arkansas, Delaware (pri­or to judi­cial abo­li­tion), Florida, Illinois (pri­or to abo­li­tion), Indiana, Louisiana, Pennsylvania, South Carolina, and Tennessee.
² Nevada, New Jersey (pri­or to abo­li­tion), Oregon, and Virginia (pri­or to abo­li­tion).
³ Texas.
⁴ Arizona, younger than age 15; Connecticut, younger than age 16 (pri­or to abo­li­tion), and Wyoming (defen­dant rea­son­ably should have known the vic­tim was younger than age 17).

Table 1. States Authorizing the Death Penalty for Killing a Child

The typ­i­cal jus­ti­fi­ca­tion for allow­ing the death penal­ty for the mur­der of a child is that chil­dren are a par­tic­u­lar­ly vul­ner­a­ble and defense­less class of vic­tims, deserv­ing of spe­cial pro­tec­tion under the law. However, as the use of this type of aggra­va­tor by oth­er states shows, that rea­son­ing is gen­er­al­ly applied with respect to infants and chil­dren, not ado­les­cents. Moreover, the set­tings in which ado­les­cents are mur­dered are mate­ri­al­ly dif­fer­ent from those in which infants and young chil­dren die. Approximately 80% of mur­der vic­tims aged 13 – 16 are killed in gun inci­dents. Among homi­cide vic­tims aged 17 – 19, that ris­es to 85%. Younger chil­dren are more vul­ner­a­ble to, and com­par­a­tive­ly more like­ly to be killed by, beat­ings, asphyx­i­a­tion, and stran­gu­la­tion: types of killings that are gen­er­al­ly regard­ed as more rep­re­hen­si­ble than most deaths by shoot­ing.⁵

In 2017, the New Hampshire and New Mexico leg­is­la­tures reject­ed death penal­ty bills that would have made the mur­der of any­one under age 18 an aggra­vat­ing cir­cum­stance.⁶ If enact­ed, those statutes would have had the broad­est age-of-vic­tim aggra­vat­ing cir­cum­stance in the United States. The U.S. Supreme Court has said that death penal­ty statutes must gen­uine­ly nar­row the offens­es for which the death penal­ty may be sought. But while the child-vic­tim aggra­vat­ing cir­cum­stances employed in most states sat­is­fy that con­sti­tu­tion­al man­date, the under-age-18 aggra­vat­ing cir­cum­stance encom­pass­es such a large per­cent­age of homi­cides that it rais­es sig­nif­i­cant constitutional questions.

According to the Bureau of Justice Statistics, ful­ly 10% of homi­cide vic­tims in the United States in the years 1980 through 2008 were under age 18.⁷ DPIC has reviewed the FBI Uniform Crime Statistics from 2010 – 2015 to esti­mate how broad­ly an expan­sion of age-of-vic­tim aggra­vat­ing cir­cum­stances to age 18 could affect eli­gi­bil­i­ty for cap­i­tal pros­e­cu­tion. As set forth in the tables below, under cur­rent homi­cide trends in the United States, an esti­mat­ed 8%- 9% of all mur­ders (between 1 in 11 and 1 in 12) would become capital offenses.

Table 2. FBI Uniform Crime Statistics (Single Victim/​Single Perpetrator Cases)

Table 3. FBI Uniform Crime Statistics (Murder Victims by Age, Expanded Homicide Data, 2010 – 2015)

FBI Uniform Crime Statistics for the years 2011 – 2015 indi­cate that 9.2% of homi­cide vic­tims killed in sin­gle victim/​single per­pe­tra­tor mur­ders were under age 18. (See Table 2.) The FBI has sort­ed these cas­es by age of per­pe­tra­tor as well, per­mit­ting an esti­mate of the per­cent­age of cas­es that would not be death eli­gi­ble because they were com­mit­ted by juve­nile offend­ers. The data indi­cate that 14% of homi­cides involv­ing vic­tims younger than age 18 — or 1.3% of all sin­gle victim/​single per­pe­tra­tor homi­cides — were com­mit­ted by offend­ers under age 18 against vic­tims under age 18. But even exclud­ing those cas­es, an esti­mat­ed 8% of all mur­ders would become cap­i­tal offens­es under an aggra­vat­ing cir­cum­stances per­mit­ting the death penal­ty for killing a vic­tim under age 18.

The FBI Uniform Crime Statistics for all homi­cides between 2010 and 2015 show that a total of 6,770 homi­cides dur­ing this peri­od (8.9%) involved vic­tims younger than age 18. (See Table 3.) The FBI does not sup­ply age of per­pe­tra­tor infor­ma­tion for these data, but assum­ing a sim­i­lar rate of offend­ing by juve­nile per­pe­tra­tors, 7.6% of all homi­cides would be death eli­gi­ble under a vic­tim-less-than-age-18 aggra­vat­ing cir­cum­stance. To under­stand the breadth of such an aggra­vat­ing cir­cum­stance, this would trans­late into 5,783 cap­i­tal-eli­gi­ble offens­es over this six-year peri­od, or rough­ly 964 new cap­i­tal offens­es per year, if adopted nationwide.

The Risks Inherent in Child-Victim Aggravating Circumstances

A 2017 study of more than 1500 cas­es in which con­vict­ed pris­on­ers were lat­er exon­er­at­ed found a direct rela­tion­ship between the per­ceived seri­ous­ness of the crime and mis­car­riages of jus­tice: ​“the ​‘worst of the worst crimes,’” the University of Denver researchers said, ​”pro­duce the ​‘worst of the worst evi­dence.’” The very same fac­tors that make child killings so hor­ri­fy­ing also make them much more sus­cep­ti­ble to wrong­ful con­vic­tion. The high­ly emo­tion­al and high­ly sen­sa­tion­al nature of these cas­es increas­es the stakes and, for the pros­e­cu­tor, the rewards of a con­vic­tion and cre­ates both con­scious and uncon­scious incen­tives for mis­con­duct. The same is true for expert wit­ness­es, and it should come as no sur­prise that a dis­pro­por­tion­ate num­ber of junk-sci­ence exon­er­a­tions and wrong­ful con­vic­tions involve the deaths of children.

One of the nation’s most famous death-row exon­er­a­tions and one of its most infa­mous wrong­ful exe­cu­tions involve the deaths of children.

Kirk Bloodsworth was con­vict­ed and sen­tenced to death in Maryland in 1984 for the bru­tal rape and mur­der of a young girl. The con­vic­tion rest­ed on faulty eye­wit­ness iden­ti­fi­ca­tions by sev­er­al child wit­ness­es and the sup­pres­sion of excul­pa­to­ry evi­dence — includ­ing a police record iden­ti­fy­ing an alter­nate sus­pect who ulti­mate­ly turned out to be the killer. Faced with this evi­dence, the jury reject­ed Mr. Bloodsworth’s ali­bi wit­ness­es. Mr. Bloodsworth was grant­ed a new tri­al, but again was con­vict­ed. This time, how­ev­er, he was sen­tenced to life. He was released in 1993 only after sub­se­quent DNA test­ing con­firmed his innocence.

Cameron Todd Willingham was con­vict­ed and sen­tenced to death in Texas in 1992 on charges that he mur­dered his three chil­dren by set­ting the house on fire. He was exe­cut­ed in 2004. His case is now wide­ly regard­ed as a wrong­ful exe­cu­tion based upon junk expert tes­ti­mo­ny. Mr. Willingham was con­vict­ed of cap­i­tal mur­der after arson inves­ti­ga­tors con­clud­ed — based upon 20 fac­tors they con­sid­ered to be indi­ca­tors of arson — that an accel­er­ant had been used to set three sep­a­rate fires inside the Willingham home. Their con­clu­sions were based upon prin­ci­ples of fire sci­ence that have since been repu­di­at­ed. Four nation­al arson experts who exam­ined the tri­al evi­dence con­clud­ed that the orig­i­nal arson inves­ti­ga­tion in the case was flawed, that the fire may well have been acci­den­tal, and that there was no sci­en­tif­ic sup­port for the prosecution’s expert tes­ti­mo­ny that the fire had been deliberately set.

Noted arson expert, the late Gerald Hurst said, ​“There’s noth­ing to sug­gest to any rea­son­able arson inves­ti­ga­tor that this was an arson fire. It was just a fire.” Former Louisiana State University fire instruc­tor Kendall Ryland added, “[It] made me sick to think this guy was exe­cut­ed based on this inves­ti­ga­tion.… They exe­cut­ed this guy and they’ve just got no idea — at least not sci­en­tif­i­cal­ly — if he set the fire, or if the fire was even intentionally set.”

As is typ­i­cal of many of the child-vic­tim exon­er­a­tion cas­es, the pros­e­cu­tion bol­stered its faulty evi­dence with false or per­jured tes­ti­mo­ny. In the Willingham case, pros­e­cu­tors pre­sent­ed false tes­ti­mo­ny from a jail­house infor­mant — a drug addict on psy­chi­atric med­ica­tion — who claimed Mr. Willingham had con­fessed to him in the coun­ty jail. Evidence dis­cov­ered years after the Willingham exe­cu­tion showed that the pros­e­cu­tion had giv­en Webb favor­able treat­ment, then delib­er­ate­ly elicit­ed per­jured tes­ti­mo­ny from Webb that he had been promised and giv­en noth­ing for his tes­ti­mo­ny.⁸

The use of junk arson sci­ence to con­demn fathers for the deaths of their chil­dren is not lim­it­ed to Texas. Similarly flawed tes­ti­mo­ny has led to the con­vic­tion of oth­er men whose chil­dren died in fires their fathers almost cer­tain­ly did not set. Here are just two examples:

Dennis Counterman was con­vict­ed and sen­tenced to death for the sup­posed arson mur­der of his two young sons. The pros­e­cu­tion whit­ed-out from a police state­ment his intel­lec­tu­al­ly dis­abled wife had giv­en them her admis­sion that she had awak­ened Mr. Counterman to let him know the house was on fire. The pros­e­cu­tion also with­held a social ser­vices record in its pos­ses­sion show­ing that one of Mr. Counterman’s sons had a his­to­ry of fire-start­ing. After win­ning a new tri­al as a result of pros­e­cu­to­r­i­al mis­con­duct and inef­fec­tive defense assis­tance, the pros­e­cu­tion threat­ened to cap­i­tal­ly repros­e­cute Counterman. To obtain his imme­di­ate release, he pled no con­test to lesser charges.

Also in Pennsylvania, Daniel Dougherty was arrest­ed four­teen years after the fatal fire that killed his two sons and was charged with their mur­der based upon an accu­sa­tion his ex-wife made dur­ing a bit­ter cus­tody dis­pute claim­ing that Mr. Dougherty had con­fessed to set­ting the fire with gaso­line. In fact, although the fire mar­shal con­duct­ed a pro­fes­sion­al­ly inad­e­quate arson inves­ti­ga­tion, inves­ti­ga­tors had test­ed for accel­er­ants and found none. To bol­ster the same type of inac­cu­rate arson tes­ti­mo­ny that was pre­sent­ed in the Willingham case, pros­e­cu­tors pre­sent­ed tes­ti­mo­ny from two prison ​“infor­mants” that Dougherty had con­fessed to them. Like the infor­mant in Willingham’s case, both of these wit­ness­es also had men­tal health prob­lems and had received undis­closed ben­e­fits for their testimony.

As a post-con­vic­tion evi­den­tiary hear­ing was about to start on the issue of tri­al counsel’s inef­fec­tive­ness in fail­ing to seek the assis­tance of an expert wit­ness in fire sci­ence, Philadelphia pros­e­cu­tors dropped the death penal­ty. The tri­al court then denied Dougherty a new tri­al, but that deci­sion was reversed by an appel­late court. However, on retri­al, the tri­al court improp­er­ly allowed the pros­e­cu­tion to read into the record the tes­ti­mo­ny of the now-unavail­able fire mar­shal who had pro­vid­ed inac­cu­rate arson tes­ti­mo­ny at Dougherty’s first tri­al. In a ​“bat­tle of the experts,” in which the pros­e­cu­tor scorn­ful­ly labeled one of the nation’s fore­most arson experts as noth­ing bet­ter than ​“a Kensington whore,” Dougherty was con­vict­ed again. That con­vic­tion, too, was over­turned. The jury in Dougherty’s third tri­al acquit­ted him of first-degree mur­der but con­vict­ed him of felony mur­der for deaths occur­ring dur­ing an arson, which in Pennsylvania car­ries a life sen­tence. Dougherty died in prison in 2021 while argu­ment on his appeal of that con­vic­tion was pending.

At least sev­en oth­er peo­ple wrong­ly con­vict­ed of child-mur­der have been exon­er­at­ed from death row. In at least three of the cas­es, there was no mur­der at all.

Vicente Figueroa Benavides was sen­tenced to death for sup­pos­ed­ly mur­der­ing his girlfriend’s 21-month-old tod­dler by rap­ing and anal­ly sodom­iz­ing her. But the girl was not raped or sodom­ized: she most prob­a­bly died of com­pli­ca­tions from hav­ing been hit by a car. In vacat­ing Benavides’s con­vic­tion in March 2018, the California Supreme Court said the foren­sic evi­dence that sent the for­mer Mexican farm­work­er to death row in 1993 was ​“exten­sive,” ​“per­va­sive,” ​“impact­ful,” and ​“false.” During oral argu­ment, Associate Justice Carol Corrigan — a for­mer pros­e­cu­tor — described the med­ical tes­ti­mo­ny describ­ing the alleged sex­u­al assault as being ​“among the most hair-rais­ing false evi­dence that I’ve encoun­tered in all the time that I’ve been look­ing at criminal cases.”

Benavides — whose lawyers have argued is devel­op­men­tal­ly dis­abled and pos­sess­es the men­tal abil­i­ty of a 7‑year-old — told the police and jury dur­ing the tri­al that he lost track of the tod­dler while he was prepar­ing din­ner on November 17, 1991 and he found her out­doors, vom­it­ing. Consuelo’s moth­er took her to a local med­ical cen­ter that evening, where her con­di­tion wors­ened. After surgery and two hos­pi­tal trans­fers, the child died a week later.

At tri­al, foren­sic pathol­o­gist Dr. James Diblin tes­ti­fied that the tod­dler had died from ​“blunt force pen­e­trat­ing injury of the anus” and claimed that the major inter­nal injuries she suf­fered were the result of rape. He fur­ther tes­ti­fied that arm injuries, inter­nal trau­ma, dilat­ed pupils, and com­pres­sion rib frac­tures that Consuelo sus­tained had been ​“caused by tight squeez­ing dur­ing a sex­u­al assault.” Dr. Jess Diamond, who eval­u­at­ed Consuelo at Kern Medical Center, also ini­tial­ly tes­ti­fied that the tod­dler had been raped. However, med­ical records obtained by Benavides’s post-con­vic­tion lawyers from California’s Habeas Corpus Resource Center (HCRC) showed that the exam­in­ing physi­cians had not seen any signs of bleed­ing when Consuelo was brought to the hos­pi­tal, and a nurse who helped treat the young girl said that nei­ther she nor any of her col­leagues saw evi­dence of anal or vagi­nal trau­ma when the child arrived. Instead, the court said, the injuries to Consuelo’s gen­i­talia and anus were ​“attribut[able] to med­ical inter­ven­tion,” includ­ing repeat­ed failed efforts to insert a catheter and the improp­er use of an adult-sized catheter on the small child.

In over­turn­ing the wrong­ful con­vic­tion, the court sug­gest­ed that the girl had nev­er been raped or sodom­ized and may not have been mur­dered at all. When he offered the opin­ion that Consuelo had been raped, Dr. Diamond had not seen either the med­ical records from her ini­tial hos­pi­tal­iza­tion or pho­tos from the autop­sy. After the HCRC lawyers showed him that evi­dence, Dr. Diamond with­drew his assess­ment that Consuelo had been raped. He wrote in an affi­davit, ​“I am con­vinced that this case presents a tremen­dous fail­ing of the crim­i­nal justice system.”

The HCRC lawyers also pre­sent­ed evi­dence from Dr. Astrid Heger, one of the country’s lead­ing experts on child abuse, who described Dr. Didbin’s asser­tion that Consuelo’s injuries had been the prod­uct of sex­u­al assault as ​“so unlike­ly to the point of being absurd. … No such mech­a­nism of injury has ever been report­ed in any lit­er­a­ture of child abuse or child assault.” She said the inter­nal injuries Consuelo sus­tained were com­mon­ly seen in vic­tims of automobile accidents.

Prosecutors admit­ted that the foren­sic evi­dence they used to con­vict Benavides was false, but asked the state court to sus­tain a con­vic­tion for sec­ond-degree mur­der. Even after agree­ing on April 17, 2018 to drop all charges, Kern County District Attorney Lisa Green refused to con­cede that Benavides was inno­cent of murder.

(Maura Dolan, California Supreme Court, cit­ing false evi­dence, over­turns mur­der con­vic­tion that put Delano man on death row, Los Angeles Times, March 12, 2018; Chloe Carlson, BREAKING NEWS: Death penal­ty reversed; ​“false evi­dence” used in tri­al, court rules, Kern Golden Empire, March 12, 2018; Jose Gaspar, With ​‘hair-rais­ing false evi­dence’ exposed, Delano man get a sec­ond chance, The Bakersfield Californian, March 9, 2018; Olivia LaVoice, Did Kern send an inno­cent man to Death Row? High court con­sid­er­ing the stun­ning pos­si­bil­i­ty, Kern Golden Empire, undated.) 

Sabrina Butler was 17 years old when she was accused of mur­der­ing her 9‑month old son. She was wrong­ly con­vict­ed and con­demned by the state of Mississippi in 1990. Butler’s baby, who had a heart mur­mur, had stopped breath­ing. After attempts at resus­ci­ta­tion failed, Butler rushed to the hos­pi­tal, where the young child was pro­nounced dead. Based on bruis­es left by her resus­ci­ta­tion attempts, she was arrest­ed the next day for alleged child abuse, inter­ro­gat­ed by the police, pros­e­cut­ed, and sen­tenced to death.

Her con­vic­tion was over­turned by the Mississippi Supreme Court in 1992, say­ing that the pros­e­cu­tion had failed to prove that the inci­dent was any­thing more than an acci­dent. Prosecutors nev­er­the­less repros­e­cut­ed her, but this time she was acquit­ted. It is now believed that the baby may have died either of cys­tic kid­ney dis­ease or from sud­den infant death syn­drome (SIDS).

Like Ms. Butler, Rodricus Crawford was con­vict­ed and sen­tenced to death for the alleged mur­der of his infant son. As with Butler, Crawford sought help after his baby was non-respon­sive. Instead, he was con­demned for mur­der in a con­tro­ver­sial death penal­ty case that had attract­ed nation­al atten­tion amid evi­dence of race dis­crim­i­na­tion and prosecutorial excess. 

According to mul­ti­ple med­ical experts, the local foren­sic pathol­o­gist botched the autop­sy and, despite evi­dence of sep­sis in the blood and pneu­mo­nia in both of the infant’s lungs, con­clud­ed that the baby had been suf­fo­cat­ed. Crawford was pros­e­cut­ed by Dale Cox, a con­tro­ver­sial Assistant District Attorney who once told the Shreveport Times that Louisiana needs to ​“kill more peo­ple” with the death penal­ty. Cox had per­son­al­ly obtained one-third of all death sen­tences imposed in Louisiana between 2010 and 2014, and had open dis­dain for black par­ents of chil­dren born out of wed­lock. In an inter­view with The New Yorker, he said he believed the ​“destruc­tion of the nuclear fam­i­ly and a tremen­dous­ly high ille­git­i­mate birth rate” had brought about an ​“epi­dem­ic of child-killings” in Caddo Parish.

Cox learned of Crawford’s case at the same time he was seek­ing the death penal­ty against Sam Jordan — a young black man with bor­der­line Intellectual Disability and Fetal Alcohol Spectrum Disorder — for the death of Jordan’s son, who had been left in his father’s care over the Christmas hol­i­day. The New Yorker reports that, when the sen­tenc­ing jury returned a life ver­dict for Jordan, ​“Cox told a local TV sta­tion, ​‘I take it as a fail­ure that I was unable to con­vince the jury to kill him.’” 

Cox uncon­sti­tu­tion­al­ly struck black jurors from Crawford’s case, which ulti­mate­ly led the Louisiana Supreme Court to over­turn Crawford’s con­vic­tion and death sen­tence. Cox then argued a venge­ful ver­sion of Scripture to the jury, quot­ing the pas­sage: ​“You shall have a mill­stone cast around your neck and you will be thrown into the sea.” By the time the Louisiana Supreme Court ruled on Crawford’s appeal, the med­ical evi­dence of inno­cence had become so pow­er­ful that the tri­al court released Crawford on bail. In April 2017, pros­e­cu­tors final­ly agreed to dis­miss all charges against him.

In 1994, in Callins v. Collins, Justice Antonin Scalia sin­gled out the bru­tal rape and mur­der of an 11-year-old girl as epit­o­miz­ing the need for cap­i­tal pun­ish­ment. ​“How envi­able a qui­et death by lethal injec­tion com­pared with that!,” he wrote. Twenty years lat­er, DNA evi­dence exon­er­at­ed Henry McCollum and his half-broth­er, Leon Brown, the intel­lec­tu­al­ly dis­abled North Carolina men who had been wrong­ly con­vict­ed and sen­tenced to death for that murder.

Madison Hobley and Kennedy Brewer also have also been exon­er­at­ed after hav­ing been wrong­ly con­demned for the deaths of chil­dren. (You can read more details about their cas­es on DPIC’s page on Innocence Cases.) The sto­ry of Mississippi prosecution’s use of false foren­sic evi­dence to con­vict Brewer and sen­tence him to death is told in detail in Radley Balko and Tucker Carrington’s 2018 book, The Cadaver King and the Country Dentist.

Child-vic­tim aggra­vat­ing cir­cum­stances put at risk oth­er par­ents or part­ners of phys­i­cal­ly frail or vul­ner­a­ble chil­dren. It is esti­mat­ed that hun­dreds of par­ents and oth­er care­givers are impris­oned, some on death row, as a result of alleged shak­en baby deaths. The junk diag­no­sis has been dis­cred­it­ed in med­ical and legal jour­nals, and yet prosecutions continue.

Ha’im Al Matin Sharif (for­mer­ly Charles Robins) was released from prison on June 7, 2017 after spend­ing more than 28 years on Nevada’s death row. He had been con­vict­ed and sen­tenced to death in December 1988 on charges that he had mur­dered his girl­friend’s 11-month-old daugh­ter, Britany Smith, in April 1988. The con­vic­tion was based upon the tes­ti­mo­ny of a local med­ical exam­in­er that Britany had been phys­i­cal­ly abused and mur­dered, and was sup­port­ed by the tes­ti­mo­ny of the vic­tim’s moth­er and oth­er rel­a­tives who tes­ti­fied to hav­ing seen Sharif abuse the baby. The jury deter­mined that Sharif had phys­i­cal­ly abused Britany over a sev­er­al-month peri­od, includ­ing break­ing her leg, and that the mur­der had been com­mit­ted by means of torture.

Mr. Sharif’s con­vic­tion and sen­tence had been upheld by the state courts on direct appeal and in ini­tial state post-con­vic­tion pro­ceed­ings. Because of a con­flict of inter­est in the Nevada Federal Defender’s office, the Federal Public Defender for the District of Arizona was appoint­ed in 2012 to han­dle Mr. Sharif’s fed­er­al court habeas cor­pus peti­tion, and for the first time in the case, his lawyers con­duct­ed a foren­sic inves­ti­ga­tion of the med­ical evi­dence relat­ing to the cause of Britany’s death and the phys­i­cal injuries she had sus­tained in the pre­ced­ing months. That inves­ti­ga­tion pro­duced evi­dence that that the baby had not been mur­dered at all, but had died from Barlow’s dis­ease (infan­tile scurvy), that Mr. Sharif had not been present at the time Britany sus­tained cer­tain oth­er injuries, and that the oth­er phys­i­cal injuries — includ­ing her bro­ken leg — were also caused by scurvy. 

As the foren­sic inves­ti­ga­tion was pro­ceed­ing, Britany’s moth­er, Lovell McDowell, and her broth­er revealed for the first time that Las Vegas police and pros­e­cu­tors had threat­ened to imprison her and take away her chil­dren unless they tes­ti­fied against Mr. Sharif. These wit­ness­es also indi­cat­ed that when they told author­i­ties that their pro­posed tes­ti­mo­ny was not true, they were instruct­ed to tes­ti­fy as they had been told to ensure that Mr. Sharif would be sen­tenced to death. Ms. McDowell specif­i­cal­ly recant­ed her tes­ti­mo­ny that Mr. Sharif had abused Britany, say­ing her tes­ti­mo­ny about abuse had been a prod­uct of threats and coer­cion. The fed­er­al court ruled that Mr. Sharif had made a suf­fi­cient show­ing that he was ​“actu­al­ly inno­cent” and that ​“his death penal­ty was improp­er­ly imposed,” and held the fed­er­al court pro­ceed­ings in abeyance to per­mit him to present the new evi­dence to the state courts.

The state tri­al court refused to hear the evi­dence, dis­miss­ing Mr. Sharif’s claims on pro­ce­dur­al grounds. On September 22, 2016, the Nevada Supreme Court unan­i­mous­ly decid­ed that Mr. Sharif ​“ha[d] pre­sent­ed spe­cif­ic fac­tu­al alle­ga­tions, that if true, would show that it is more like­ly than not that no rea­son­able juror would have con­vict­ed him of first-degree mur­der and child abuse beyond a rea­son­able doubt or found the sin­gle aggra­vat­ing cir­cum­stance used to make him death eli­gi­ble.” It ordered the low­er court to con­duct an evi­den­tiary hear­ing on his inno­cence claim and on the claims of police and pros­e­cu­to­r­i­al mis­con­duct. Subsequently a doc­tor retained by the Clark County District Attorney’s office agreed that Britany had suf­fered from scurvy. Although pros­e­cu­tors would not agree to a com­plete exon­er­a­tion, they offered to vacate Mr. Sharif’s con­vic­tion and death sen­tence for first-degree mur­der and amend the judg­ment to sec­ond-degree mur­der with a sen­tence of time served, which would result in Mr. Sharif’s immediate release.

Jeffrey Havard (pic­tured) was con­vict­ed and sen­tenced to death in Mississippi in 2002 for the alleged mur­der of his girlfriend’s 6‑month-old daugh­ter. The key evi­dence against him came from Dr. Steven Hayne, a med­ical exam­in­er whose prac­tices and diag­noses have been harsh­ly crit­i­cized by experts in foren­sic pathol­o­gy. Hayne tes­ti­fied at Havard’s tri­al that he found symp­toms of ​“shak­en baby syn­drome” and sex­u­al abuse on the infant. The sex­u­al abuse find­ing — which Hayne lat­er dis­avowed — allowed pros­e­cu­tors to seek the death penal­ty in the case. 

Dr. Hayne acknowl­edged hav­ing per­formed as many as 1,700 autop­sies in a year — far more than the num­ber rec­om­mend­ed by experts — in addi­tion to hav­ing his own pathol­o­gy prac­tice. According to Dr. David Fowler, chief med­ical exam­in­er in Maryland and a for­mer chair­man of the stan­dards com­mit­tee for the National Association of Medical Examiners, that num­ber is ​“beyond defen­si­ble.” Hayne’s autop­sy in Havard’s case has been reviewed and dis­cred­it­ed by oth­er expert pathol­o­gists. Other cas­es in which he had pro­vid­ed piv­otal pros­e­cu­tion tes­ti­mo­ny have been over­turned through DNA test­ing. Investigations into Hayne’s cre­den­tials indi­cate he had nev­er been cer­ti­fied in foren­sic pathol­o­gy by the American Board of Pathology. He had tak­en the exam in 1980, but failed it.

In April 2015, the Mississippi Supreme Court unan­i­mous­ly ruled that Havard was enti­tled to an evi­den­tiary hear­ing to present new sci­en­tif­ic evi­dence on the unre­li­a­bil­i­ty of shak­en baby diag­noses. At the time, Dr. Hayne admit­ted to The Jackson Clarion-Ledger news­pa­per that there was ​“grow­ing evi­dence” that his shak­en baby diag­no­sis was ​“prob­a­bly not cor­rect” and that he ​“did­n’t see any evi­dence of sex­u­al assault.” In that evi­den­tiary hear­ing in August 2017, Dr. Hayne backed off his shak­en baby diag­no­sis, but still insist­ed the death was homi­cide. The court did not per­mit tes­ti­mo­ny on his recan­ta­tion of his sex­u­al abuse diag­no­sis. Defense expert, Dr. Michael Baden, tes­ti­fied that ​“shak­ing had noth­ing to do with the death. All the injuries,” he said, ​“were entire­ly con­sis­tent with blunt force impact, con­sis­tent with­in the man­ner in which Mr. Havard on day one said that the baby had fallen.” 

The evi­den­tiary hear­ing con­clud­ed in August 2017. More than a year lat­er, in September 2018, Adams County Circuit Judge Forrest Johnson over­turned Havard’s death sen­tence but upheld his con­vic­tion. Forrest wrote that Hayne’s recan­ta­tion of his shak­en-baby diag­no­sis was ​“not suf­fi­cient to under­mine this court’s con­fi­dence in the con­vic­tion,” but ​“there is a cau­tious dis­tur­bance in con­fi­dence of the sen­tence of death, even if slight.” Prosecutors sub­se­quent­ly dropped the death penal­ty against Havard, who con­tin­ues to appeal his conviction.

(R. Balko, The mur­der evi­dence evap­o­rat­ed, but Jeffrey Havard still sits in a Mississippi prison, September 19, 2018; R. Balko, The Watch, Controversial med­ical exam­in­er backs off ​‘shak­en baby’ claim in death penal­ty case, The Washington Post, August 16, 2017; C. Madden, How did baby Britt die? Shaken baby syn­drome at cen­ter of hear­ing in 2002 death penal­ty case, The Nachez Democrat, August 15, 2017; J. Mitchell, Changing sci­ence may give death row inmate new hear­ing, The Jackson Clarion-Ledger, April 2, 2015; R. Balko, Despite Evidence From Discreted Medical Examiner, Mississippi’s Jeffrey Havard Nears Execution, Huffington Post, January 8, 2012; C. Robertson, Questions Left for Mississippi Over Doctor’s Autopsies, New York Times, January 7, 2013.)

Elizabeth and Samuel Glick were Amish dairy farm­ers. They took their 4‑month-old daugh­ter, Sara Lynn, to the hos­pi­tal, where she died. Doctors saw hem­or­rhag­ing in her right eye and exten­sive bruis­ing, and sus­pect­ed child abuse; the autop­sy showed blood and swelling in her brain. The coun­ty coro­ner ruled her death a homi­cide. Parents like the Glicks face poten­tial cap­i­tal pros­e­cu­tion under child-vic­tim aggra­vat­ing cir­cum­stances. But there was no homi­cide. The actu­al cause of Sara Lynn’s death was a com­bi­na­tion of vit­a­min K defi­cien­cy and rare genet­ic (bile­salt trans­porter) disorder.

Vulnerable or dis­fa­vored defen­dants are also at increased risk. In the case of Commonwealth v. Alejandro Mendez in Centre County, Pennsylvania, near Penn State University, a Costa Rican immi­grant was cap­i­tal­ly charged and held in cus­tody for two years with­out a tri­al after the death of his infant son, Lucas, from sub­dur­al bleed­ing and reti­nal hem­or­rhag­ing. Again, there was no crime. The Mendez fam­i­ly had come to the United States so his wife could be treat­ed for can­cer. However, at the same time she was breast feed­ing Lucas, she had been giv­en antibi­otics that imped­ed the baby’s absorp­tion of vit­a­min K. As a result, Lucas devel­oped a bleed­ing dis­or­der and could not clot. Prosecutors would not admit their error, and — fac­ing death — Mendez was forced to plead no con­test to manslaugh­ter charges and be deport­ed to Costa Rica.

⁵ See, e.g., FBI Uniform Crime Statistics, Crime in the United States 2011, Table 9: Murder Victims by Age by Weapon, https://ucr.fbi.gov/crime-in-the‑u.s/2011/crime-in-the‑u.s.- 2011/ta­bles/­ex­pand­ed-homi­cide-data-table‑9.

⁶ HB 351. Although the pre­ca­to­ry lan­guage to House Bill 351 described the pro­pos­al as ​“AN ACT mak­ing a per­son who know­ing­ly caus­es the death of a child guilty of cap­i­tal mur­der,” the bill actu­al­ly would have autho­rized cap­i­tal pros­e­cu­tion of a defen­dant when­ev­er ​“he know­ing­ly caus­es the death of … [a]nother who is less than 18 years of age.”

⁷ Alexia Cooper and Erica L. Smith, Bureau of Justice Statistics, Homicide Trends in the United States, 1980 – 2008 (Nov. 2011), https://​www​.bjs​.gov/​c​o​n​t​e​n​t​/​p​u​b​/​p​d​f​/​h​t​u​s​8​0​0​8.pdf.

⁸ See DPIC, Executed But Possibly Innocent. See also Steve Mills and Maurice Possley, Texas Man Executed on Disproved Forensics, Chicago Tribune (December 9, 2004); Anderson Cooper, Was an Innocent Man Executed in Texas?, Anderson Cooper 360 Blog (April 9, 2007); Maurice Possley, The Prosecutor and the Snitch. Did Texas exe­cute an inno­cent man?, The Marshall Project (August 3, 2014).

By Robert Brett Dunham
August 16, 2017

The sec­tion on Risks Inherent in Child-Victim Aggravating Circumstances was updat­ed on April 18, 2018 to include dis­cus­sion of the Benavides, Brown, and McCollum cas­es and on August 30, 2020 to include devel­op­ments in the Jeffrey Havard case. It was fur­ther updat­ed January 25, 2022 to include devel­op­ments in Daniel Dougherty’s case. Other sec­tions were updat­ed as indi­cat­ed in the text.