USCA8 Reverses Decision that Denied Justice Because It Was Delayed
The U.S. Court of Appeals for the Eighth Circuit today reinstated the death sentence of Missouri double murdered Carman Deck. The court reversed a district judge decision that had overturned Deck’s sentence because he had to be resentenced a third time, effectively holding that justice has to be denied altogether because it had been badly delayed.
The Crime
The facts of this case are described in the Missouri Supreme Court’s original opinion, State v. Deck, 994 S.W.2d 527 (1999):
In June 1996, Deck planned a burglary with his mother’s boyfriend, Jim Boliek, to help Boliek obtain money for a trip to Oklahoma. Deck targeted James and Zelma Long, the victims in this case, because he had known the Longs’ grandson and had accompanied him to the Longs’ home in DeSoto, Missouri, where the grandson had stolen money from a safe. The original plan was to break into the Longs’ home on a Sunday while the Longs were at church. In preparation for the burglary, Deck and Boliek drove to DeSoto several times to canvass the area.
On Monday, July 8, 1996, Boliek told Deck that he and Deck’s mother wanted to leave for Oklahoma on Friday, and he gave Deck his .22 caliber High Standard automatic loading pistol. That Monday evening, Deck and his sister, Tonia Cummings, drove in her car to rural Jefferson County, near DeSoto, and parked on a back road, waiting for nightfall. Around nine o’clock, Deck and Cummings pulled into the Longs’ driveway.
Deck and Cummings knocked on the door and Zelma Long answered. Deck asked for directions to Laguana Palma, whereupon Mrs. Long invited them into the house. As she explained the directions and as Mr. Long wrote them down, Deck walked toward the front door and pulled the pistol from his waistband. He then turned around and ordered the Longs to go lie face down on their bed, and they complied without a struggle.
*532 Next, Deck told Mr. Long to open the safe, but because he did not know the combination, Mrs. Long opened it instead. She gave Deck the papers and jewelry inside and then told Deck she had two hundred dollars in her purse in the kitchen. Deck sent her into the kitchen and she brought the money back to him. Mr. Long then told Deck that a canister on top of the television contained money, so Deck took the canister, as well. Hoping to avoid harm, Mr. Long even offered to write a check.
Deck again ordered the Longs to lie on their stomachs on the bed, with their faces to the side. For ten minutes or so, while the Longs begged for their lives, Deck stood at the foot of the bed trying to decide what to do. Cummings, who had been a lookout at the front door, decided time was running short and ran out the door to the car. Deck put the gun to Mr. Long’s head and fired twice into his temple, just above his ear and just behind his forehead. Then Deck put the gun to Mrs. Long’s head and shot her twice, once in the back of the head and once above the ear. Both of the Longs died from the gunshots.
Three juries — 36 people — have unanimously decided that death is the appropriate punishment for this crime. It certainly is, but it has been a long time coming.
The Delays
The Missouri Supreme Court affirmed on direct appeal. It considered and rejected a claim that Deck’s rights had been violated by his lawyer’s omission from a jury instruction of the paragraphs that expressly informed the jury that they did not have to be unanimous when finding mitigating circumstance, unlike the requirement for aggravating circumstances. That instruction was intended to implement the U.S. Supreme Court’s decision in Mills v. Maryland, 486 U.S. 367 (1988), a dubious decision that many Justices of the high court have indicated they believe to be incorrect. The Missouri court held that when the instructions were read as a whole they conveyed that mitigation findings need not be unanimous.
Even so, when the case returned to state supreme court on collateral review, they overturned the judgment for ineffective assistance of counsel in making the error they had rejected as a ground of relief before. The reason for the turnabout ostensibly was that the standard for reviewing claims under the ineffective assistance standard of Strickland v. Washington, 466 U.S. 668 (1984) is lower than the plain-error standard on direct review. That seems doubtful to me, though I don’t claim to be knowledgeable on Missouri’s plain error standard.
Having won a dubious reversal in his favor, Deck was sentenced again. The second jury found, unanimously, six aggravating factors: Deck committed the murders while engaged in the commission of another unlawful homicide; Deck murdered each victim for the purpose of pecuniary gain; each murder involved depravity of mind; each murder was committed for the purpose of avoiding a lawful arrest; each murder was committed while Deck was engaged in a burglary; and each murder was committed while Deck was engaged in a robbery. Not surprisingly the second jury came to the same conclusion as the first. These crimes warrant the death penalty. This time the case went to the United States Supreme Court.
In second sentencing proceeding, Deck had been in shackles in front of the jury. Supreme Court precedent has long limited restraint of defendants in front of jury during the guilt phase of trial. In Holbrook v. Flynn, 475 U.S. 560 (1968) the court referred to shackling the defendant in front of the jury as an “inherently prejudicial practice that … should be permitted only where justified by an essential state interest specific to each trial.” What is and is not such a practice is determined by the inferences that jurors are likely to draw from it.
In the guilt phase of trial, the defendant’s appearance in shackles may lead the jurors to believe that the defendant is particularly dangerous, further leading to an inference that he is guilty of the crime, the very issue they are called upon to decide. This may interfere with the presumption of innocence the defendant is entitled to at this point.
What about at the penalty phase? There presumption of innocence is gone. The guilt verdict finds it rebutted beyond a reasonable doubt. What inference might the jury draw from seeing the defendant in shackles that they would not draw from the fact that he murdered two helpless old people in cold blood? As Justice Thomas noted in dissent, “Treating shackling at sentencing as inherently prejudicial ignores the commonsense distinction between a defendant who stands accused and a defendant who stands convicted.” See Deck v. Missouri, 544 U.S. 622, 651 (2005).
Nonetheless, the majority did extend the shackling rule to the penalty phase, and Deck won the lottery a second time.
The Re-re-sentencing
At the third penalty trial, Deck’s lawyer did not object that the time spent on death row violated the Eighth Amendment. Such claims have been kicking around for a long time. They are often called “Lackey claims” after Justice Stevens’ dissenting opinion in Lackey v. Texas, 514 U.S. 1045 (1995). The Supreme Court has declined to take this claim up every time it has been raised, and no more than two Justices at a time have ever said it warrants consideration.
Deck’s attorney also did not raise a more focused claim that a third sentencing proceeding conducted so many years after the crime was prejudicial to his client. Indeed, time is usually in the defendant’s favor in such matters. Long-ago crimes do not generally stir the same kind of outrage as recent shocking events.
Counsel did have trouble finding anyone to testify in Deck’s favor, though. Counsel testified that one relative he contacted “had provided favorable testimony at the two earlier penalty trials” but she was “now doing a 180.” Mitigation in capital cases often involves a lot of dubious testimony from family members, and evidently this one had gotten fed up with providing it.
Nonetheless, the testimony of the now-unavailable or now-uncooperative witnesses still existed, either on video or on transcript, and defense counsel could and did introduce it. The testimony was obviously not compelling the first or second times, as the juries found death to be warranted anyway, and it remained not compelling the third time.
The Present Case
Deck’s lawyers did not make the delay claim on direct appeal or on state collateral review. They made a lot of other claims, of course, all of which were rejected.
In federal court on habeas corpus, Deck made 32 claims. Thirty were rejected by the district court. Two were claims that he had never presented to the state courts. As the district court stated the first one, “Deck claims that the inordinate delay between his conviction … and his final sentencing trial … violated his rights to due process and to be free from the infliction of cruel and unusual punishment.” The second claim was that trial counsel’s failure to make the first claim was ineffective assistance in violation of the Sixth Amendment.
The district court was doubly wrong. First, the court made up a new rule of law and imposed it retroactively on the states. The Supreme Court held over thirty years ago in Teague v. Lane, 489 U.S. 288 (1989) that federal courts sitting in habeas corpus cannot do that. Although the state objected on this ground, the district judge did not even mention Teague in her opinion. That is inexcusable. To overturn a judgment in a capital case with no discussion at all of a landmark Supreme Court precedent that appears to preclude that action and has been briefed by the opposing party is conduct unbecoming a judge.
The only statement that even comes close to a Teague discussion is the district judge’s passing comment that “[a]t the time Deck’s sentencing judgment became final, the law was clearly established that, in a capital case, a criminal petitioner has a constitutionally protected right to provide the jury with mitigating evidence.” This is followed by a citation to a case that is not a Teague case but instead relates to the rule regarding claims that have been presented to the state court and rejected by them on the merits.
CJLF’s amicus curiae brief on behalf of the Long family in the court of appeals notes:
If that statement of a clearly established “rule” at a very high level of generality was intended to dispose of the Teague issue, it did not come close to the depth of analysis required.
From the beginning of the Teague line of cases decades ago through the most recent decisions, the Supreme Court has repeatedly and emphatically rejected the argument that a decision can be exempted from the “new rule” limitation merely by citing a broad principle that the new rule is deemed to support. The very next year after Teague, the high court noted in Sawyer v. Smith, 497 U.S. 227, 236 (1990), “the test would be meaningless if applied at this level of generality.” The high court has repeated this holding many times since. See, e.g., Gray v. Netherland, 518 U.S. 152, 169 (1996). “We have repeatedly told courts . . . not to define clearly established law at high level of generality.” Ashcroft v. al-Kidd, 563 U.S. 731, 742 (2011) (citing Sawyer).
The second error involves the requirement imposed by law that state convicts take their claims to the state courts first before bringing them to the federal courts. See 28 U.S.C. § 2254(b). That rule has exceptions, but the district judge’s claim to have found one does not hold water, the court of appeals held today.
One exception to the rule is that if the attorney in state court provided ineffective assistance in violation of the defendant’s Sixth Amendment rights and the claim is no longer available in state court, the federal court may proceed to the merits anyway. In a state court proceeding where the defendant has no federal constitutional right to counsel, including state collateral review, this exception generally does not apply. See Coleman v. Thompson, 501 U.S. 722, 755 (2012).
In Martinez v. Ryan, 566 U.S. 1, 9 (2012), the Supreme Court created what it called a “narrow exception” to the Coleman rule. When the state collateral review lawyer fails to make a claim of ineffective assistance of the trial lawyer under circumstance that amount to the collateral review lawyer also being ineffective, that will provide sufficient cause for the federal court to proceed to the merits. The court of appeals described the district court’s holding on this point:
The district court held that Deck’s case fell squarely under the Martinez exception. In its view, the defaulted claim was substantial, because counsel at Deck’s third penalty-phase trial had a difficult time mounting a mitigation case because of the passage of time. Specifically, some of Deck’s witnesses from the first two penalty-phase trials were unable or unwilling to provide in-person testimony the third time around. In light of this difficulty, the court believed it was obvious that counsel should have raised Eighth and Fourteenth Amendment objections before the trial began. It was so obvious, in fact, that postconviction counsel was ineffective for failing to recognize it later.
The district court then took the cause analysis one step further. Relying on Edwards v. Carpenter, 529 U.S. 446 (2000), it concluded that the newly excused ineffective-assistance-of-trial-counsel claim provided cause for the default of the underlying Eighth and Fourteenth Amendment claims. See id. at 453 (leaving open the possibility that a petitioner who can overcome the default of an ineffective-assistance claim can use it to show cause for the default of another claim). This Martinez-plus-Edwards approach allowed the court to grant habeas relief for both the ineffective-assistance-of-trial-counsel claim and the underlying constitutional claims, even though Deck never raised them in state court.
But that is all wrong.
Every step in this analysis, however, still depends on getting through the Martinez gateway first. This means that the key question is whether postconviction counsel was ineffective. If not, there is no excuse for the failure to raise trial counsel’s ineffectiveness during state postconviction proceedings. See Martinez, 566 U.S. at 14. And if trial counsel’s ineffectiveness is defaulted without excuse, then so are Deck’s underlying Eighth and Fourteenth Amendment claims. See Edwards, 529 U.S. at 452–53. Without Martinez, Deck never gets to Edwards.
Far too many judges seem to think that any lawyer who omitted any claim that they subsequently find has merit was per se ineffective. That is not the law. Throwing everything conceivable against the wall to see if any of it sticks is not effective advocacy. The Supreme Court recognized in Smith v. Murray, 477 U.S. 527, 536 (1986), that the “process of ‘winnowing out weaker arguments on appeal and focusing on’ those more likely to prevail, far from being evidence of incompetence, is the hallmark of effective appellate advocacy.”
The Lackey claim and related claims are way out there on a limb. There is no precedent to support them. Other claims are likely to appear better candidates to an attorney who is doing exactly what the Supreme Court said he should do.
As we have explained, failing to make an argument that would “require the resolution of unsettled legal questions” is generally not “outside the wide range of professionally competent assistance.” Id. at 836 (quotation marks omitted); see Parker v. Bowersox, 188 F.3d 923, 929 (8th Cir. 1999) (explaining that it is not objectively unreasonable for counsel to “fail[] to anticipate a change in the law”). When postconviction counsel filed Deck’s petition in 2010, the law was far from settled that a 10-year delay between conviction and sentencing would give rise to a constitutional claim, much less that trial counsel was ineffective for failing to raise the argument two years earlier.3 See New v. United States, 652 F.3d 949, 953 (8th Cir. 2011) (concluding that the absence of “controlling authority” supporting a legal argument doomed an ineffective-assistance claim).
Footnote 3 at this point notes the court’s doubt about the whole claim at its very root.
3 We further note that it is doubtful that Deck has made a “substantial” claim that he was prejudiced, even if trial counsel’s performance had been objectively unreasonable. He suggests that the passage of time deprived him of mitigating evidence, including from certain witnesses who were unavailable to testify at his third trial. Even so, we doubt that there is “a reasonable probability that” the trial court would have done anything different had it faced a timely objection from Deck’s trial counsel. Strickland, 466 U.S. at 694. After all, much of the missing testimony was cumulative to other evidence or did not add much to Deck’s mitigation case.
In the end:
In sum, postconviction counsel’s performance was reasonable. It follows that the Martinez exception—the only conceivable basis for excusing Deck’s procedural default—is unavailable to him.
So the death sentence is reinstated. Deck will, I expect seek further review of this decision. However, this is the last appeal he gets as of right, and I can’t see either the Eighth Circuit en banc or the Supreme Court taking this up.
And Missouri is one of the states that remains willing and able to enforce its death penalty. Long overdue justice may come in 2021.
The case is Deck v. Jennings, No. 17-2055.
CJLF’s amicus curiae brief on behalf of the Long family, by Kym Stapleton, is available here.

Is there any way to calculate the amount of Missouri taxpayer resources that have been spent on this case over the last quarter century? I am not trying to troll, but sincerely wondering what the price of capital justice might be in this (still pending) case.
Not really. Genuine efforts to calculate the cost of capital litigation have been frustrated by inadequate data. The estimates most commonly cited are inflated ones by anti-death-penalty advocates more interested in providing ammunition for their side’s arguments than in finding the truth. We had quite a battle over that in California’s initiative wars of four and eight years ago.
What is clear, though, is that a large chunk of the cost is unnecessary. The initial reversal was based on Mills v. Maryland, an exemplar of the illegitimate “tinkering with the machinery of death” cases denounced in my last law review article (in a journal I think you are familiar with).
Thanks for the response. I share your concerns about inadequate data in this space, including data on costs saved by capital charges leading to quick pleas (which is an addition area for your research). Do you think the current federal Administration might be more willing than most to work on this needed research?
I have no information on the thinking of the current Administration on this subject.
No matter who is in charge, I’d like to see the feds try to collect some data on these issues so that it is not simply subject to claims by advocates.
Quality data would be good to have, but that is much easier said than done. Throughout criminal statistics, we see lots of info that is easy to gather and less info on more important matters that are more difficult to gather and put into a national database.
Agreed, which is why we have long needed national leadership on this front for a range of criminal justice issues — e.g., both supporters and opponents of marijuana reform ought to support the feds collecting a lot more detailed data on the CJ impact of these state reforms. Indeed, I would like to see the feds have a much smaller role in operating CJ systems (e.g., they should not prosecute so many of the local cases that can and should be done by states), and a much larger role in collecting and analyzing data about these systems.