To understand why Idaho built a room full of gun ports, you have to understand one man’s veins, and you have to understand that his veins were never really the point.
Thomas Eugene Creech was born September 9, 1950.1 By the time Idaho police linked him to a killing in 1974, he already had a history that would eventually make him one of the most written-about prisoners in the state’s history, and one of the hardest to fit into any tidy account of guilt, punishment, and proportion. Creech was convicted of two murders committed that year and initially sentenced to death, a sentence a court reduced to life imprisonment on appeal two years later.2 He did not stay off death row. In 1981, while already imprisoned, Creech beat a fellow inmate, David Dale Jensen, to death in what prosecutors characterized as an unprovoked attack inside the prison itself.3 He was sentenced to death for that killing in 1983, and he has been under that sentence ever since, making him the longest-serving death row prisoner in Idaho’s history and among the longest-serving in the nation.
The scale of what Creech may have done beyond his convictions is genuinely difficult to state responsibly, because the record is a mixture of solid evidence and claims that were never corroborated. Over the years, Creech confessed at various points to as many as forty-two killings across multiple states, some tied, in his own telling, to an unnamed biker gang and drug-running network, and at least one confession invoking a satanic church.4 Almost all of that is uncorroborated. What investigators have been able to substantiate is narrower and still serious: police have linked him with reasonable confidence to somewhere between nine and eleven killings, and he has been convicted in connection with five.5 As recently as January 2024, a cold-case investigation by the San Bernardino County Sheriff’s Department in California concluded that Creech was responsible for the 1974 murder of a man named Daniel Walker, decades after the fact.6 I recount this not to diminish the horror of what Creech did, which is real and substantial by even the most conservative accounting, but because his case sits at the extreme edge of what the American death penalty is built to handle, and extreme cases have a way of revealing what a system actually values when it is tested.
The Clemency Vote That Split Down the Middle
On January 19, 2024, the Idaho Commission of Pardons and Parole heard Creech’s clemency petition. Ten days later, on January 29, the board announced that it had deadlocked three to three.7 Under the state’s rules, a tie is not a win for the person seeking mercy. It is treated as a failure to reach the majority required to recommend clemency, and so the deadlock functioned, procedurally, exactly as a denial would have. One detail from that hearing deserves more attention than it has generally received: one of the original prosecutors from Creech’s first trial, a man named Jim Harris, said afterward that he had come to believe Creech should be taken off death row, not out of sympathy for what Creech had done but because, in his assessment, the case had already cost Idaho taxpayers an enormous sum chasing an execution that might, in his words, never actually happen.8 This is worth sitting with. A prosecutor who helped put Creech on death row in the first place had concluded, decades later, that the machinery built to carry out his sentence had become its own kind of failure, independent of any question about Creech’s guilt.
Governor Little, who holds final authority over clemency in Idaho’s capital cases, declined to override the board’s deadlock. Judge Jason D. Scott signed a new death warrant the following day, setting execution for February 28, 2024.9
The Room Where Nothing Worked
What happened on the morning of February 28, 2024 is the hinge of this book, and it deserves to be taken slowly, because the details matter and because the state’s own after-action language repays scrutiny sentence by sentence.
Creech was moved into F-Block, the same unit that now houses the firing squad chamber, in the standard pre-execution process. On the morning of February 28, the execution team began the process of establishing intravenous access for the lethal injection protocol. Over the course of roughly an hour, according to reporting from the scene and the department’s own subsequent statements, the team made eight separate attempts to establish a usable line. All eight failed.10 Around eleven in the morning, IDOC Director Josh Tewalt made the decision to stand down the execution rather than continue attempting access or fall back on some improvised alternative. Creech, who had been prepared for death, restrained, and presumably had spent that hour believing this was the day he would die, was instead unstrapped and returned to a cell. The death warrant expired at midnight that night without having been carried out.
Tewalt’s statement to the press afterward is a document worth reading in full, because it is the clearest early example of a rhetorical pattern this entire book tracks: the reframing of a visible failure as evidence of institutional integrity. “While the execution of Thomas Eugene Creech did not proceed, the process worked as intended,” he said. “Our medical team experienced difficulty establishing an IV site of sufficient quality to carry out the execution by lethal injection. It’s a scenario for which we trained and prepared, and I made the decision to stand down the execution based on the professional advice of our medical team. It was the right decision. Some are characterizing today’s events as a failure, but the opposite is true. The process worked to prevent a failure, and I think that’s an important distinction.”11
Read that passage again and notice what it is actually claiming. It is not claiming that the execution succeeded. It is claiming that the failure to execute was itself a species of success, because the alternative, an execution botched in progress rather than aborted beforehand, would have been worse. This is not an unreasonable position to hold about that specific morning. Standing down rather than forcing a ninth or tenth attempt at a vein was almost certainly the right call in the narrow sense that Tewalt meant it. But notice what the statement does not do. It does not ask why the state was in that position at all. It does not ask whether an institution that requires eight failed venous attempts before choosing not to continue has actually demonstrated competence, or whether it has demonstrated exactly the opposite, dressed in the language of professionalism and dignity. “Infused this difficult process with dignity and respect,” Tewalt said of his staff’s conduct that morning.12 I do not doubt that the individuals involved conducted themselves with as much composure as the circumstances allowed. But dignity in the execution of a botched procedure is not the same thing as the procedure having been dignified, and this book will keep returning to that distinction because Idaho’s entire pivot to the firing squad depends on readers not noticing it.
What a Death Warrant Day Looks Like
It is worth describing, briefly, what an actual execution date does to the ordinary functioning of the institution around it, because the bureaucratic texture of the day makes the abstraction of “carrying out a sentence” considerably more concrete. On the day Idaho attempted to execute Thomas Creech in February 2024, the department reduced speed limits on the roads surrounding the prison complex, canceled ordinary visiting hours at every facility south of Boise, and opened designated areas for demonstrators, all standard operating procedure for a scheduled execution rather than anything specific to Creech’s case.13 Media witnesses were briefed on logistics and timelines well in advance and escorted, as a group, from an administrative holding area to the viewing room at a scheduled hour, a choreography practiced enough that the department could describe it to reporters in advance with confidence, even though the execution itself did not, in the end, proceed.
This matters not because traffic patterns or visitor schedules bear directly on this book’s argument, but because the sheer bureaucratic ordinariness of the surrounding machinery is itself part of what this book has been describing throughout. An execution, in the state’s own operational planning, is treated as a logistics problem alongside speed limits and visitor management, planned with the same institutional competence a state applies to any other scheduled public function. That competence is real, and it is also beside the point. The state can coordinate traffic control and visitor schedules with total reliability while still failing, as this chapter has just described, at the one task all of that coordination exists to support.
The Second Attempt, and the Argument That Stopped It
The state did not let the matter rest. On October 16, 2024, IDOC served Creech with a new death warrant, scheduling his execution for November 13. This time, the department announced, it had changed its protocol: rather than relying solely on peripheral veins in the arms, the execution team would be authorized to use deep central venous access, meaning sites in the neck, chest, or groin, if the earlier sites again proved inaccessible.14
Creech’s attorneys at the Federal Defender Services of Idaho argued that subjecting a man to a second execution attempt after the state had already failed once amounted to cruel and unusual punishment under the Eighth Amendment, independent of any question about the method’s theoretical reliability going forward.15 Their public statement at the time framed the stakes starkly: “Mr. Creech has spent more than 50 years in prison and is now suffering from significant mental health issues because of the trauma he was subjected to when the state failed to execute him. We hope the courts will recognize the cruel and unusual level of punishment that this remorseful and harmless old man has already been through, and stop a needless execution.”16
The phrase “harmless old man” deserves flagging without either endorsement or dismissal, because it captures something true about the peculiar cruelty of extreme longevity on death row that this book returns to later. Whatever Creech did in 1974 and 1981, the man who sat in F-Block in November 2024 was seventy-four years old and had spent, by that point, close to five decades in some stage of capital proceedings. Punishment stretched across that much time does something to the moral texture of a sentence that the original jury, sentencing a man in his twenties or thirties, could not have anticipated and did not vote on.
The execution did not happen, though not on the ground Creech’s attorneys had pressed hardest. On November 6, 2024, a week before the scheduled date, U.S. District Judge G. Murray Snow of Arizona, sitting by designation, stayed the execution — not because he had found merit in the Eighth Amendment claim, but to give the parties time to brief a separate set of claims that prosecutors had acted improperly at Creech’s clemency hearing, and to await the Idaho Supreme Court’s ruling on the state-law versions of the double jeopardy and cruel-and-unusual arguments. “I’m not going to rush justice in a case like this,” Snow said. The death warrant expired on its own terms.17 As of this writing, in July 2026, Creech remains alive on Idaho’s death row, his case unresolved, his age now seventy-five, the state’s oldest and most visible failure of capital process and, not coincidentally, the single case legislators cited most often while building the argument for House Bill 37.
Three weeks later the state court answered, and it answered against Creech. On November 27, 2024, the Idaho Supreme Court unanimously held that a second execution attempt would violate neither the Double Jeopardy Clause nor the Eighth Amendment, reasoning that the psychological strain and pain Creech described were inherent in any method of execution, and that the first attempt had not completed the punishment the legislature authorized. I do not think that reasoning survives contact with the account this chapter has just given. But it is the law in Idaho as of this writing, and any reader weighing what stands between Thomas Creech and the chair in F-Block should know that his strongest constitutional argument has already been heard by his state’s highest court and rejected without a dissent.
The Prosecutor’s Later Doubt, Revisited
Jim Harris’s reflection, noted earlier, carries a fuller significance than it first appears to. Harris was not a defense attorney, an abolitionist advocate, or an outside critic of Idaho’s capital punishment system. He was one of the prosecutors who built the case that first sent Thomas Creech to death row, a person whose professional life was invested in securing exactly the outcome Idaho spent five decades failing to carry out. When a person in that position concludes, decades later, that the pursuit of an execution has become its own kind of failure independent of the underlying question of guilt, that conclusion carries a different evidentiary weight than the same argument made by someone who opposed the death penalty from the outset.
I raise this because it previews a pattern this book will return to in its final chapters: the people best positioned to recognize the machinery of capital punishment as dysfunctional are often the people who have spent the most time operating it, not outside critics who arrived at their conclusions from principle alone. Josh Tewalt’s own statement after the failed 2024 execution, discussed earlier in this chapter, shows an institutional actor straining to describe a failure as a success. Jim Harris’s statement, from a different vantage point inside the same system, shows an institutional actor arriving, without any apparent ideological agenda, at something closer to this book’s own diagnosis: that Idaho’s pursuit of Creech’s execution had become, independent of any judgment about his guilt, a project whose costs had stopped making sense on its own terms.
A Pattern, Not an Anomaly
Idaho’s failure with Creech did not happen in a vacuum, and it is important to resist the temptation, which Idaho’s own legislative debate indulged constantly, to treat it as a uniquely Idahoan embarrassment rather than a symptom of a national collapse. Lethal injection has been failing publicly, in state after state, for close to two decades. Oklahoma’s execution of Clayton Lockett in 2014 became a national story when a poorly placed IV allowed drugs to leak into surrounding tissue rather than his bloodstream, and Lockett writhed and spoke for over twenty minutes before dying of a heart attack roughly forty minutes after the drugs began, rather than by the intended drug protocol.18 Alabama’s attempted execution of Doyle Hamm in 2018 was called off after roughly two and a half hours of failed attempts to find usable veins, leaving Hamm’s attorney to describe a body covered in puncture wounds.19 Alabama tried again with Alan Miller in 2022 and called that one off too, for the same underlying reason: no usable vein, found in time, before the death warrant expired.20 Kenneth Smith, whose 2024 nitrogen execution I witnessed and describe later in this book, had himself survived a failed lethal injection attempt in Alabama in 2022, the state having given up after more than an hour of trying to establish access before his warrant lapsed.21
I list these cases together because Idaho’s legislature, in the floor debate over House Bill 37, treated Creech’s case as though it were an isolated engineering problem, solvable with a better protocol or a different execution method. It was not isolated. It was the local instance of a pattern playing out across nearly every state that still relies on lethal injection, a pattern with a specific, well-documented cause: the same pharmaceutical industry that manufactures the drugs states want to use for killing has spent the last fifteen years trying to keep those drugs out of execution chambers, refusing to sell to corrections departments, tightening distribution agreements, and forcing states into a black market of compounding pharmacies and expired stock that produces exactly the kind of unreliable, botched procedures Creech experienced.
The firing squad’s appeal to legislators, across every state that has adopted or expanded it in the past several years, is precisely that it does not depend on a supply chain any manufacturer can choke off. A rifle and a bullet require no pharmaceutical company’s cooperation. This is, from the state’s perspective, the method’s core engineering advantage, and it is worth naming plainly: Idaho did not turn to the firing squad because the evidence showed it was more humane. It turned to the firing squad because the evidence showed lethal injection was becoming impossible to reliably supply, and the firing squad solves a supply-chain problem, not a suffering problem.
The Clemency Board’s Silence
One further detail about the Idaho Commission of Pardons and Parole’s three-to-three deadlock deserves attention, because the board’s own reasoning, or the absence of any detailed public reasoning, is itself informative. Clemency hearings of this kind typically produce some public record of the considerations weighed, whether in the form of published opinions, dissenting statements, or at minimum a summary of the arguments presented. The three-to-three deadlock on Creech’s petition does not appear to have generated the kind of detailed public reasoning that would let outside observers understand precisely what divided the board, whether the disagreement centered on questions of guilt, questions of proportionality given Creech’s age and decades of incarceration, or questions of process given the seventh board member’s absence that Creech’s subsequent appeal specifically challenged.
This matters for the same reason the shield laws matter: the more opaque a decision-making process is, the harder it becomes for the public to hold the institution accountable for the reasoning behind its decisions, or to learn from that reasoning when the next difficult case arises. A clemency board that explains its reasoning in detail, even where that reasoning ultimately supports denial, gives future boards, future legislators, and the public a record to build on. A board that deadlocks without full public explanation leaves only the bare vote count, three to three, denied, for history to work with, a pattern of institutional opacity this book has now traced from the clemency board to the shield laws protecting Idaho’s shooters to the state’s own initial public statements after Creech’s failed execution.
What the State Learned, and What It Refused to Learn
Here is the distinction this chapter leaves you with. Idaho’s Department of Correction, faced with the failure in F-Block on February 28, 2024, drew a narrow and technical lesson: peripheral IV access is unreliable in some subset of cases, so the state should either change its venous access technique or change its method entirely. That is the lesson reflected in both the deep-central-vein protocol change attempted for the November 2024 date and, more decisively, in House Bill 37 itself.
There is a different lesson the state could have drawn…and did not. An institution that requires eight failed attempts at a man’s veins before calling off his execution, that then tries again with a more invasive protocol on the same man eight months later, that builds a room with gun ports rather than reconsider whether it should be attempting to kill people at all, has not learned that killing is hard to do reliably. It has learned that reliability is the only value at stake. Chapter six shows exactly how reliable Idaho’s chosen replacement actually is, using the fresh, well-documented record South Carolina produced between Idaho’s vote and the day that law took effect. The short answer, which the rest of this book will support in detail, is: not reliable enough to justify anything the legislature said about it on the floor in March 2025.
The Longest-Serving Man in the Country
I noted earlier that Creech is the longest-serving death row prisoner in Idaho’s history. It is worth being clear about how that ranks nationally. Creech’s time under his current death sentence, imposed in 1983 for the killing of a fellow inmate, runs to more than forty years as of this writing, and if his earlier 1970s death sentence is counted as part of a continuous, if legally interrupted, relationship to Idaho’s machinery of capital punishment, the figure runs closer to fifty.22 Nationally, prisoners who have spent more than four decades under sentence of death are rare but not unique to Idaho; several states, California and Florida prominent among them, hold prisoners whose sentences predate the birth of some of the corrections officers now assigned to guard them. What distinguishes Creech’s case is not merely the length of the wait but the fact that Idaho actually attempted, and failed, to end it, leaving him in the singular position of a man who has now survived his own execution and continues, at seventy-five, to wait for a second attempt under an entirely different method than the one that failed the first time.
What the Federal Defenders Meant by “Trauma”
One phrase from Creech’s attorneys’ 2024 statement deserves more careful unpacking than a single quotation can provide: their claim that Creech was “suffering from significant mental health issues because of the trauma he was subjected to when the state failed to execute him.” This is not merely a rhetorical flourish aimed at generating sympathy. Clinical literature on prisoners who have survived execution attempts, though limited by the rarity of the phenomenon, describes a specific and recognizable psychological injury: the experience of being prepared for death, restrained, and made to believe with total sincerity that one’s life is about to end within minutes, followed by an abrupt and unexplained reprieve, produces symptoms that overlap substantially with acute trauma responses documented in survivors of other near-death experiences, hypervigilance, intrusive recollection, a diminished capacity to trust any subsequent assurance that a threat has passed.23 Creech’s attorneys were not describing an abstraction. They were describing a man who had been walked, by the state’s own hour-long account, to the edge of his death and then walked back, and who now lives with the knowledge that the state intends, at some undetermined future date, to walk him there again, this time with a rifle rather than a needle waiting at the end of the corridor.
Idaho spent 1.2 million dollars solving the wrong problem. Thomas Creech, and the seven other people on the state’s row, are the ones who will find out what that mistake actually costs.
Notes
011. Associated Press (Rebecca Boone), contemporaneous reporting on Thomas Creech, 2023–2024; Death Penalty Information Center, “Idaho,” state-by-state profile.
022. “Thomas Creech: Idaho execution fails after IV line couldn’t be established,” CNN, February 28, 2024; Death Penalty Information Center, “Clemency Request for 73-Year-Old Death Row Prisoner in Idaho Has Support of Trial Judge and Prosecutor,” March 14, 2025. Creech’s sentencing history for the Jensen killing runs through three dates and is reported inconsistently as a result. He was first sentenced to death in 1981; that sentence was set aside and he was resentenced to death in January 1983, the date the Idaho Department of Correction’s own roster carries; and he was resentenced again in 1995 following federal habeas relief, which is why some court records, including the Idaho Supreme Court’s 2024 opinion in Creech v. State, describe him as “sentenced to death in 1995.” This book uses 1983 throughout, following IDOC’s roster, and treats the 1995 proceeding as a resentencing on the same conviction rather than a new sentence.
033. “Idaho delays execution of serial killer Thomas Creech after failed lethal injection attempts,” CBS News, February 28, 2024 (citing the Ada County Prosecutor’s Office); “Judge postpones execution of serial killer Thomas Creech months after he survived a botched execution attempt,” CNN, November 9, 2024.
044. “Commission of Pardons and Parole denies Creech a commutation recommendation,” Idaho Reports, January 29, 2024 (quoting the U.S. Supreme Court’s 1993 findings on the number of Creech’s victims); “Serial Killer Thomas Eugene Creech Execution Halted After IV Failure,” Forbes, February 28, 2024.
055. “Commission of Pardons and Parole denies Creech a commutation recommendation,” Idaho Reports, January 29, 2024 (evidence linking Creech to eleven deaths; five murder convictions).
066. “Idaho parole board denies Thomas Creech’s petition for clemency hearing,” KTVB, October 29, 2024 (San Bernardino County cold-case investigation into the 1974 murder of Daniel Walker); “Thomas Creech, Idaho’s longest serving death row inmate denied clemency,” KMVT/KSVT, January 30, 2024.
077. “Commission of Pardons and Parole denies Creech a commutation recommendation,” Idaho Reports, January 29, 2024; “Idaho Commission of Pardons and Parole denies death row inmate’s commutation recommendation,” Idaho Capital Sun, January 30, 2024.
088. “‘There has been a real loss of confidence and support’: Experts weigh in on death penalty in Idaho,” KTVB, April 25, 2024 (quoting former prosecutor Jim Harris); see also Death Penalty Information Center, “Idaho Parole Board Grants Clemency Hearing for 73-Year-Old Death Row Prisoner Facing Execution,” March 14, 2025 (quoting Harris).
099. “Idaho Commission of Pardons and Parole denies death row inmate’s commutation recommendation,” Idaho Capital Sun, January 30, 2024; Idaho Department of Correction, “Execution updates,” idoc.idaho.gov.
1010. Idaho Department of Correction, “Execution updates,” idoc.idaho.gov; “Thomas Creech: Idaho execution fails after IV line couldn’t be established,” CNN, February 28, 2024.
1111. Idaho Department of Correction, “Execution updates,” idoc.idaho.gov (statement of Director Josh Tewalt, February 28, 2024).
1212. IDOC, “Execution updates.”
1313. Idaho Department of Correction, “Execution updates,” idoc.idaho.gov (operational notices ahead of the February 28, 2024 execution date).
1414. Idaho Department of Correction, “IDOC Serves Death Warrant to Thomas Creech,” October 16, 2024; “State of Idaho Schedules New Date for the Execution of Thomas Creech,” KPVI, October 16, 2024.
1515. “New death warrant served for Thomas Creech after failed execution in February,” KTVB, October 16, 2024.
1616. KTVB, October 16, 2024 (quoting the statement of the Federal Defender Services of Idaho).
1717. “Judge postpones execution of serial killer Thomas Creech months after he survived a botched execution attempt,” CNN, November 9, 2024; Kevin Fixler, “Idaho Supreme Court denies death row prisoner’s appeal. Federal stay of execution ordered,” Idaho Statesman, November 6, 2024; Death Penalty Information Center, “Idaho: Federal Judge Grants Stay of Execution for Thomas Creech,” 2024 (Judge G. Murray Snow’s remarks at the November 5 hearing and his written stay of November 6). On the Idaho Supreme Court’s subsequent ruling, see Creech v. State, Docket No. 52327 (Idaho, November 27, 2024) (holding that a second execution attempt violates neither the Double Jeopardy Clause nor the Eighth Amendment).
1818. This account of the 2014 Clayton Lockett execution in Oklahoma reflects contemporaneous Associated Press reporting.
1919. This account of the 2018 Doyle Hamm execution attempt in Alabama reflects contemporaneous Associated Press reporting.
2020. This account of the 2022 Alan Miller execution attempt in Alabama reflects contemporaneous Associated Press reporting.
2121. This account of Kenneth Smith’s 2022 execution attempt in Alabama reflects contemporaneous Associated Press reporting; see also the discussion of Smith’s subsequent 2024 nitrogen execution in the Introduction.
2222. Death Penalty Information Center, “Clemency Request for 73-Year-Old Death Row Prisoner in Idaho Has Support of Trial Judge and Prosecutor,” March 14, 2025 (“sentenced to death in 1981”).
2323. General clinical literature on acute trauma responses in survivors of near-death experiences; see also “New death warrant served for Thomas Creech after failed execution in February,” KTVB, October 16, 2024 (Federal Defender Services of Idaho statement on Creech’s mental-health deterioration).