On Point

7th Circuit decision; on habeas review of Wis. opinion No. 2005AP2599-CR

Habeas – Filing Deadline – Tolling

The District Court had authority to grant Socha’s pre-filing, pre-deadline request to extend the 28 U.S.C. § 2244(d) deadline for his habeas petition, made on the ground of equitable tolling.

… First, there is no absolute bar imposed by Article III on judicial actions closely connected with a case or controversy that has not yet been filed. Perhaps the best- known example of a court’s taking action with respect to a case that has yet to be filed comes in Federal Rule of Civil Procedure 27, which permits depositions to perpetuate testimony. …

Second, it is possible to view a motion like Socha’s as the actual petition for a writ of habeas corpus, filed in an incomplete form but with a promise to furnish supporting documentation later. See, e.g., Smith v. Barry, 502 U.S. 244, 248-49 (1992) (treating appellate brief as notice of appeal); Listenbee v. City of Milwaukee, 976 F.2d 348, 350-51 (7th Cir. 1992) (treating motion for extension of time as notice of appeal). No one disputes the fact that his petition would have been timely if July 15, 2008, was the date on which it was effectively filed.

… Thus, Socha’s ancillary proceeding satisfied the traditional standing requirements of injury-in-fact, causation, and redressibility. Elk Grove Unified School Dist. v. Newdow, 542 U.S. 1, 11-12 (2004).

Socha filed his request one day before the deadline would have run. He argued that because he was in segregation the prison severely limited his law library access, and he argued that the deadline should therefore be extended on the basis of equitable tolling. A district judge agreed, and ordered that the deadline be extended. Socha filed his petition within the extended deadline, but the judge then assigned to the case ruled that the first judge had no authority to order the extension, and the petition was accordingly dismissed as outside the limitation period. The 7th Circuit, as the block quote indicates, holds to the contrary. The remedy isn’t to reinstate the petition outright but, rather, to remand for a closer look. One possibility is that, as suggested above, Socha’s pre-deadline request itself be taken as a timely habeas petition; this is for the district court to determine on remand (the 7th Circuit doesn’t say why it can’t make this call, or what factors should inform the issue on remand). Another possibility is that the extension was justified on the basis of equitable tolling (or estoppel):

… Socha informed the court on several occasions that he was trying, or had tried, to file on time, but he was hindered by the adverse party (the warden). Even before the Supreme Court’s decision in Holland, we had not closed the door on the possible applicability of doctrines such as equitable tolling and equitable estoppel. See, e.g., Tucker v. Kingston, 538 F.3d 732, 734 (7th Cir. 2008). In Holland, the Court confirmed that equitable tolling is available for cases governed by § 2244(d)’s filing periods. 130 S. Ct. at 2562 (explaining that a habeas corpus petitioner is “entitled to equitable tolling only if he shows (1) that he has been pursuing his rights diligently, and (2) that some extraordinary circumstance stood in his way and prevented timely filing”) (internal quotation marks and citations omitted). We see no reason why, in an appropriate case, equitable estoppel would not also be available.

The court stresses that Socha did file his petition before undertaking the appeal. (“We thus have no need to decide whether we might have entertained an appeal had Socha filed one before he presented his petition to the district court, and we are free to reach the difficult characterization question that this case presents.”) Otherwise, had the appeal merely been before the court on a pre-petition extension ruling, the procedural situation would be much different. The court strongly suggests, albeit without distinctly holding, that in that situation it would follow United States v. Leo, 203 F.3d 162, 164 (2d Cir. 2000) (“Here, because Leon has not yet filed an actual § 2255 petition, there is no case or controversy to be heard, and any opinion we were to render on the timeliness issue would be merely advisory. Accordingly, we lack jurisdiction to consider the issue on appeal.”).

State v. Christopher Melendrez, 2009AP2070, District 4, 9/2/10

court of appeals decision (3-judge, not recommended for publication); for Melendrez: David R. Karpe; BiC; Resp.; Reply

SVP – Retroactivity of Qualifying Offense Legislation

Third-degree sexual assault wasn’t an SVP-qualifying offense when Melendrez plea-bargained a reduction of 2nd-degree sexual assault to 3rd. But by the time he was released from prison, the legislature had passed 2005 Wis. Act 434, which amended § 980.01(6)(a) to include 3rd-degree sexual assault as a “sexually violent offense,” thereby subjecting Melendrez to ch. 980 commitment. He argues that applying the amended legislation to him had a retroactive effect that violated due process.

Because it is well-established that ch. 980 isn’t penal in nature, and therefore doesn’t violate ex post facto and related principles, Melendrez is relegated to arguing due process protection against retroactive legislation in the civil context, Landgraf v. USI Film Products, 511 U.S. 244 (1994). But the statute must first have retroactive effect before due process analysis is triggered.

¶23      Wisconsin case law also establishes that retroactive legislation does not necessarily violate the due process clause. In Barbara B. v. Dorian H., 2005 WI 6, ¶¶19-20, 277 Wis. 2d 378, 690 N.W.2d 849, the court explains that a court is first to determine if a statute actually has a retroactive effect, which turns on whether it affects a substantive right that vested or accrued before the enactment. If the court concludes that it does have a retroactive effect, the court then determines whether the retroactive effect comports with due process.[11]

The court deems Act 434 non-retroactive, and therefore doesn’t need to reach due process. The Act is no more retroactive, which is to say not all all, than ch. 980 itself, which upon promulgation gathered within its reach individuals already convicted of qualifying offenses:

¶28      Melendrez also attempts to distinguish his situation from the persons who became subject to the Kansas statute—or to WIS. STAT. ch. 980—when it was first enacted.[13] He recognizes that those persons would not have known about the statute when they engaged in conduct and entered into pleas subsequently defined as sexually violent offenses. His situation is different, he asserts, because the existence of ch. 980 when he committed the offense and entered into the plea to third-degree sexual assault made it reasonable to rely on the terms of the statute at that time.[14] Rephrasing this argument reveals its flaw: Melendrez is asserting that he had a reasonable expectation that the statute would not be amended even though persons subject to ch. 980 when it was first enacted did not have a reasonable expectation that no such statue would be enacted in the first instance. Beyond asserting that there is a difference, Melendrez does not explain why his situation creates a vested interest in the version of the statute in effect when he pled—and, thus, makes the amendment retroactive as to him—when the original enactment did not have a retroactive effect on the persons first subject to ch. 980. Moreover, Melendrez’s attempt to distinguish on this ground is inconsistent with State v. Tabor, 2005 WI App 107, 282 Wis. 2d 768, 699 N.W.2d 663.

Is this a fair comparison? A pre-980 defendant couldn’t possibly have had a vested expectation (or right) of non-commitment, for the simple reason that commitment didn’t exist. Defendants wouldn’t have accepted or rejected pleas in the expectation of avoiding something — exposure to commitment — that wasn’t in place. But Melendrez, in contrast, likely would have plea-bargained the charge to 3rd- from 2nd-degree precisely to avoid commitment, which by contrast was a concrete possibility. (Or so, for present purposes, it might be assumed, though to be sure the record appears to be silent on the point, ¶28 n. 14. The idea is that regarding the two contexts as indistinguishable is much too facile, and in this sense it is fair to at least hypothesize that Melendrez’s plea-based conviction was induced by a reasonable expectation he was beyond the reach of an extant commitment regime.) As for Tabor, which upheld a reduced burden of proof at trials on petitions commenced after the statutory amendment’s effective date, ¶¶29-33: again, it is not the same thing. The court simply avoids discussing whether inducing a conviction with an assurance of no possibility of commitment — again, assuming that to be so and recognizing that we don’t know the facts — creates a vested right (or expectation) of no commitment.

Does Melendrez have any recourse against the conviction itself? That question, of course, wasn’t before the court on appeal solely of the commitment. Start with the assumption that the possibility of a 980 commitment is a mere collateral consequence of a plea, therefore doesn’t support (post-sentencing) plea withdrawal, State v. Myers, 199 Wis. 2d 391, 544 N.W.2d 609 (Ct. App. 1996). Consider, though, that the very distinction between collateral and direct consequences has been thrown into doubt by Padilla v. Kentucky. And then take into account the possibility that Melendrez was in fact informed that he would not be subject to commitment under 3rd-degree and that that is why he pleaded to the reduced charge — even under pre-Padilla caselaw you stood a chance of plea-withdrawal if you were “actively misinformed” on a collateral consequence, e.g., State v. Koll, 2008AP1403, 4/8/09. Melendrez might not have been “actively misinformed,” but if he was assured he couldn’t be committed on this plea then he might as well have been. Not an easy argument to make, but it is there.

State’s Waiver

The State waived its right to argue that Melendrez’s post-commitment motion was untimely, by failing to raise it in response to his motion: “We decline to address the State’s argument on untimeliness because it was not raised in the circuit court. State v. Champion, 2008 WI App 5, ¶17, 307 Wis. 2d 232, 744 N.W.2d 889 (‘We generally do not review an issue raised for the first time on appeal.’),” ¶13 n. 6.

And the waived argument? That 980 being civil, a motion for new 980 trial on the basis of newly discovered evidence must be brought under § 806.07 and, therefore, brought within one year of verdict.

Newly Discovered Evidence – Re-normed Actuarial

Post-trial release of new norms for Static-99, with reduced recidivism rates, didn’t support new trial, because it isn’t reasonably probable the verdict would have been different with this new information:

¶43      First, as the circuit court explained, Dr. Rosell’s testimony criticized the Static-99 on the same grounds set forth in the article, although in a less detailed and comprehensive way. The jury was informed that the original Static-99 overstated the risk of recidivism, although they did not learn by how much.

¶44      Second, the State’s experts used actuarial instruments, besides the Static-99, which the article does not address. The jury heard Dr. Rosell’s criticism of those instruments.

¶45      Third, the State’s experts acknowledged the limitations of all the actuarial tools in determining the risk of a particular person reoffending and both experts relied on much other information in forming their opinions. Specifically, one or both relied on Melendrez’s testimony of sexual offenses, his reoffending after treatment, his poor adjustment to supervision, and his lack of insight into the gravity of his offenses.

Test for newly discovered evidence recited at ¶36; court notes uncertainty in caselaw as to whether reasonable-probability portion of test is reviewed de novo or deferentially, but deems the question unnecessary to resolve in present case, ¶36 n. 17.

State v. Jennette L. Ellifritz, 2010AP713-CR, District 2, 9/1/10

court of appeals decision (1-judge, not for publication); for Ellifritz: Gary Grass; BiC; Resp.; Reply

Obstructing – Unanimity – Course of Conduct

Because Ellifritz’s actions occurred during a single course of action, over a short (40-second) period of time, instructional failure to require agreement as to which separate act constituted obstructing didn’t violate her right to unanimous verdict; State v. Giwosky, 109 Wis. 2d 446, 326 N.W.2d 232 (1982) followed, and State v. Crowley, 143 Wis. 2d 324, 422 N.W.2d 847 (1988) distinguished.

Modest disagreement will be registered here with the court’s analysis, though not the result. Giwosky indeed appears to be controlling. Giwosky, during a single altercation, both punched and threw a log at the victim and juror unanimity wasn’t required as to which act established the charge of battery. The court correctly says that Crowley can’t be applied here, but for somewhat a different reason than posited by the court. Crowley, according to the court, holds that due process is violated where it’s unclear which of two different “modes of proof” the verdict rests on, ¶15. That’s part of the story. More specifically, Crowley holds that “where the jury may have arrived at its verdict by one of two independent grounds,” the evidence must be sufficient on both, else the verdict must be set aside, 143 Wis. 2d at 334-35. In Crowley’s own instance, the evidence supported both modes, so the verdict was upheld. Crowley, then, has only a glancing relevance to Ellifritz, at best. If one of her acts relied on by the State (through, for example, closing argument) to prove obstructing were deemed legally insufficient, then Crowley would enter the picture. But that doesn’t seem to be among her arguments.

Speaking, though, of Crowley: the problem discussed there — a general verdict is supportable on one ground but not another and it’s impossible to tell which the jury relied on in assigning guilt — was known as a “Yates” issue. At the time, the error was considered  “structural,” but that is no longer true: an invalid theory embedded in a general verdict now  is subject to subject to harmless error analysis, Hedgpeth v. Pulido, 129 S.Ct. 530 (2008). Just thought you’d like to know.

Obstructing – Sufficiency of Proof

Evidence to support obstructing held sufficient, without regard for whether investigation actually hindered:

¶19      Ellifritz contends that Hamilton created a rule that a defendant’s conduct must hinder an officer’s investigation in some way before the conduct amounts to obstruction. We reject that contention. Hamilton merely stands for the proposition that refusing to provide one’s name to an officer is not obstruction per se. Hamilton did not judicially create an element not found in the statute—that the officer must have been “hindered” in his or her investigation.  In fact, this court has held that an obstruction conviction will be upheld if the jury finds that the defendant made the officer’s job “more difficult.” Grobstick, 200 Wis. 2d at 249-50. In this case, we have testimony from three officers that their investigation was made moredifficult than it should have been because of Ellifritz’s actions. Thus, to adopt Ellifritz’s argument would be to extend Hamilton beyond the holding of that decision.

The court separately indicates that Ellifritz’s false statement to the police — that marks on a 2-year-old’s face were caused by bug bites rather than having been struck — “alone would seem to amount to obstruction,” ¶5 n. 3, citing State v. Caldwell, 154 Wis. 2d 683, 686, 454 N.W.2d 13 (Ct. App. 1990) (“knowingly giving false information with intent to mislead constitutes an obstruction as a matter of law.”). However, the State doesn’t raise this argument so the court doesn’t actually reach it.

State v. Charles G. Jury, 2010AP622-CR, District 2, 9/1/10

court of appeals decision (1-judge, not for publication); for Jury: Eric R. Pangburn; BiC; Resp.

Reasonable suspicion supported stop of vehicle for any or all of the following reasons: dim tail light; necklace hanging from rearview mirror so as to obstruct driver’s view; driving on double yellow line.

City of Oshkosh v. Richard A. Selquist, 2010AP862, District 2, 9/1/10

court of appeals decision (1-judge, not for publication); for Selquist: Walter Arthur Piel, Jr.; BiC; Resp.; Reply

The police had reasonable suspicion to continue temporary detention of Selquist and to request filed sobriety testing while investigating a traffic accident:

¶7        …  In reviewing whether the officer’s further investigation and request for field sobriety tests were warranted, we apply the same standard as for an initial stop. State v. Betow, 226 Wis. 2d 90, 94-95, 593 N.W.2d 499 (Ct. App. 1999). …

…

¶9        Selquist asserts that the officer lacked reasonable suspicion that he was impaired based onWilson’s acknowledgment that upon initial contact he did not find Selquist’s behavior suspicious. Selquist further contends that, even with individual contact, Wilson’s suspicions of intoxication did not rise to the level required under the Fourth Amendment. However, Selquist’s argument ignores the circumstances under whichWilson was conducting his investigation. Initially, Wilson was trying to clear the mounting traffic at the scene. Thus, he had only momentary contact with all parties to the accident for the limited purpose of determining whether they were injured and whether their cars were drivable before instructing them to relocate to a less trafficked area for further investigation. Given the circumstances, it was not unreasonable for the officer to have failed to notice that Selquist was intoxicated. However, upon talking to him individually in a less distracted environment, Wilson noticed that Selquist had the odor of intoxicants on his breath and was slurring his speech, not to mention Selquist admitted to consuming alcoholic beverages and could not recall when he had stopped.

State v. Adamm D.J. Linton, 2010 WI App 129; for Linton: Joseph E. Redding; BiC; Resp.; Reply

Interrogation – Ambiguous Request for Counsel

Initial custodial questioning terminated when Linton invoked his right to silence. During subsequent re-interrogation, Linton said, “when I asked for a lawyer earlier, why wasn’t he appointed to me?” The detective indicated that if Linton was asking for a lawyer then the police would “just stop talking to” him. Linton agreed to talk. The trial court found that Linton had not in fact requested counsel during the first interrogation, a finding not challenged and therefore taken as given on appeal. Operative principles are settled:

¶8 The right to counsel is invoked when a suspect expresses a “‘desire to deal with the police only through counsel.’” State v. Jones, 192 Wis. 2d 78, 94, 532 N.W.2d 79 (1995) (citation omitted). Such a statement must be unambiguous—in other words, the suspect “must articulate his desire to have counsel present sufficiently clearly that a reasonable police officer in the circumstances would understand the statement to be a request for an attorney.” Davis v. United States, 512 U.S. 452, 459 (1994). If the suspect makes an ambiguous or equivocal reference to an attorney, officers need not stop questioning the suspect and may clarify the comment. Id.; see State v. Ward, 2009 WI 60, ¶43, 318 Wis. 2d 301, 767 N.W.2d 236.

Applying these principles, the court of appeals agrees with the trial court “that Linton voluntarily waived his right to an attorney when the detective sought clarification in response to Linton’s ambiguous request and Linton agreed to proceed with the interrogation,” ¶11. The court doesn’t quite explain its conclusion, perhaps because it’s self-evident: given that Linton did not in fact previously request counsel, his reference to this non-event necessarily made the detective quizzical about whether Linton indeed was now asserting his right to counsel.

Note that this was a pre-charge interrogation, meaning that 5th, rather than 6th, amendment principles apply. As to the latter: State v. Forbush, currently pending in the supreme court, raises the issue of whether an ambiguous invocation of counsel during 6th (and as opposed to 5th) amendment interrogation requires cessation of questioning.

An aside: you might wonder why the police were allowed to re-interrogate after Linton clearly invoked his right to silence. Turns out he clammed up only because “he did not like these female detectives, … they were somewhat disrespectful and aggressive, he didn’t like their manner and therefore did not want to speak to them,” ¶11. Linton doesn’t raise the point, so the court doesn’t discuss it, but we can assume that his right to silence was scrupulously honored by substituting genteel male for ill-mannered female detectives. Snark aside, it is surely true that the idea Linton knew enough to stiff-arm a couple of detectives he didn’t like strongly suggested that he knew he could put an end to the tag-teaming — had he wanted to.

Joinder/Severance

Joinder of two different homicides, 6 days apart and involving different weapons and co-actors, was proper because of an overlap in evidence:

¶14 We independently examine the propriety of the initial determination of joinder as a matter of law.  State v. Locke, 177 Wis. 2d 590, 596, 502 N.W.2d 891 (Ct. App. 1993). “The joinder statute is to be construed broadly in favor of initial joinder.” State v. Hoffman, 106 Wis. 2d 185, 208, 316 N.W.2d 143 (Ct. App. 1982). Joinder may be obtained when two or more crimes “are of the same or similar character.” Wis. Stat. § 971.12(1). “To be of the ‘same or similar character’ … crimes must be the same type of offenses occurring over a relatively short period of time and the evidence as to each must overlap.” State v. Hamm, 146 Wis. 2d 130, 138, 430 N.W.2d 584 (Ct. App. 1988).

…

¶16      We first determine whether the initial joinder was appropriate. Linton contends that joinder was improper because there was an insufficient overlap in evidence. He points out that the only overlapping evidence was the testimony of the medical examiner who conducted the Cuey autopsy and who read the notes prepared by the medical examiner who conducted the England autopsy[4] and the “scant” testimony of the detective who took Morris’s statement connecting Linton to the England homicide.[5]

¶17      In light of the broad construction we afford the joinder statute in favor of joinder, see Hoffman, 106 Wis. 2d at 208, we conclude that the aforementioned overlap in testimony is sufficient because both cases involved homicides that ensued after efforts were made to take property from another (England, attempted armed robbery; Cuey, armed burglary) within an approximately one-week time frame. As further support for this conclusion, we note that if the cases against Linton had been tried separately, testimony would have been introduced connecting the two homicides because it was only through police investigation of the Cuey homicide that they connected Linton to the England homicide. Thus, we agree with the State’s assessment that “the manner in which the England death investigation focused on [Linton], the fact that he was implicated in the Cuey murder was admissible in the England trial to explain to the jury how police identified Linton as England’s assailant.”

Not so much missing altogether but swallowed up by the court’s deference to the joinder decision, the following principle: “because the evidence of each crime would be admissible in separate trials for each, joinder was proper under Wis. Stat. § 971.12(1),” State v. Bruce T. Davis, 2006 WI App 23, ¶14. There, the court granted relief against joinder, concluding that the m.o.’s were different, with no common scheme, and one count likely wouldn’t have been admissible on separate trial of the other. (Same panel, incidentally, that decided this case.) Focus, then, is on admissibility at separate trials: if the evidence would come in anyway, it’s cost-effective to try them together. The State, for its part, argued that the investigation of Cuey’s death led to Morris, who fingered Linton on England’s homicide: “Overlap of the witnesses presented on each offense can justify joinder. United States v. Hang Le-Thy Tran, 433 F.3d 472, 478 (6th Cir. 2006).” That case upholds joinder “where the evidence of those counts is intertwined” (further characterized by that court as “significant overlap of the witnesses presented on each offense”). The Linton court bought the argument with little if any embellishment. But here’s the rub: Morris didn’t testify and his statement came in solely through the detective’s testimony; a classic Bruton problem. Maybe there were good tactical reasons not to object on confrontation grounds or, for that matter, not to renew the misjoinder/severance objection. Maybe there weren’t. The court points out: “Linton ‘concede[d] that Mr. Morris could provide admissible testimony at both trials.’ Morris, however, did not ultimately testify at Linton’s trial,” ¶16 n. 5. So you can see the basis for thinking at some point there was some evidentiary overlap. But in the end there was none, not in the sense of admissible testimony anyway, and at no time was there significant overlap.

Now, for what may be a bit of a tangent. It oughtn’t be blithely assumed that the jury is necessarily entitled to know just why the spotlight turned on the defendant. “In contexts other than a photographic identification, the phrase ‘based on information received’ may be used by police officers to explain their actions, but only if necessary to rebut a suggestion that they acted arbitrarily and only if the use of that phrase does not create an inference that the defendant has been implicated in a crime by some unknown person,” State v. Branch, 182 N.J. 338, 865 A.2d 673 (2005). In other words, this is a highly fact-contingent problem, with admissibility as a last rather than first resort. This path is also strewn with potential confrontation hazards. U.S. v. Maher, 454 F.3d 15, 23 (1st Cir 2006) (“The dividing line often will not be clear between what is true background to explain police conduct (and thus an exception to the hearsay rule and thus an exception to Crawford) and what is an attempt to evade Crawford and the normal restrictions on hearsay.”). Making the same sort of point: U.S. v. Silva, 380 F.3d 1018, 1020 (7th Cir. 2004); U.S. v. Price, 458 F.3d 205, 210 (3rd Cir 2006) (“courts must not allow the government, in the guise of ‘background explanation’ testimony by police officers, to put before the jury eyewitness accounts of bad acts by the defendant that the jury would not otherwise have heard”). Whether these sorts of problems actualized themselves in Linton’s trial can’t be ascertained, but the careful litigator might want to be sensitive generally to this potentially recurrent problem.

The other instances of “overlap” — one medical examiner read the notes of another examiner in a separate autopsy; both involved homicides and (attempted) theft — seem insubstantial (with the first example raising a potential confrontation problem as well). One count alleged that the victim was shot on a street by one of two people. The other, that the victim died from blunt force trauma to the head from bolt cutters, during a burglary participated in by 3 individuals, ¶¶2-3. Some “overlap.”

Evidence – Autopsy Photos

Autopsy photographs were properly admitted into evidence, on the felony-murder count:

¶28 As explained by the trial court, if the jury determined that Linton was guilty of felony murder as a party to the crime, the verdict form directed it to determine, as a penalty enhancer, whether Linton committed the underlying crime of burglary while armed with a dangerous weapon. The statutory definition of a “dangerous weapon” includes any “instrumentality which, in the manner it is used or intended to be used, is calculated or likely to produce death or great bodily harm.” Wis. Stat. § 939.22(10).  Bolt cutters are not generally considered a dangerous weapon but can be as used here. After reviewing the photographs, we conclude that the trial court properly exercised its discretion in admitting them into evidence. The trial court was sensitive to the nature of the photos, mindful of the potential prejudice and concluded that this was outweighed by their probative value. Under our deferential standard of review, we cannot say that the trial court erroneously exercised its discretion when it allowed the photographs to be presented during Linton’s trial. Cf. Sage v. State, 87 Wis. 2d 783, 790, 275 N.W.2d 705 (1979).

John M. Stephenson v. Levenhagen, 7th Cir No. 09-2924, 08/26/2010

7th Cir decision; petition for rehearing denied 1/14/11, 3 dissents from denial of en banc review

Habeas – Effective Assistance – Stun Belt

Counsel’s failure to object to placement of stun belt on Stephenson during trial was held by the state court to be deficient: accepting that conclusion (albeit with apparent reluctance), the federal court holds on habeas review that the deficiency wasn’t prejudicial.

Lots more to it than that — authorship by Judge Posner alone makes reading the opinion worthwhile — brief mention will be made here only of prejudice. Wisconsin caselaw strongly suggests that preserved error on shackling (or other restraint) is necessarily prejudicial. Thus, State v. Tatum, 191 Wis. 2d 547, 553, 530 N.W.2d 407 (Ct. App. 1995): “Prejudice was inherent in the cases on direct appeal in which a defendant or defense witness appeared in shackles without an extreme need for them. But ineffective assistance of counsel cases have developed a specific definition of prejudice which is to be used in those cases.” And State v. Kevin M. Champlain, 2008 WI App 5, ¶28 n. 9 (same, re: taser). This line of cases appears to mean that by establishing error you establish a violation of due process; the State then must prove harmlessness beyond reasonable doubt. But failure to register contemporaneous objection throws the defendant into ineffective-assistance territory, where the burden shifts to you. As Stephenson illustrates (and the just-cited Wisconsin cases underscore), the client is much better served by contemporaneous objection.

The dissent from denial of en banc review elaborates on this point:

 The Supreme Court’s jurisprudence makes clear that  imposing a visible  restraint on the accused is inherently prejudicial to his right to a fair trial. … I believe their  analysis overlooks  the inherent, unquantifiable prejudice of a visible restraint and is otherwise  inconsistent with the Supreme Court’s decisions on this subject.

…

… Visible restraints have been deemed “a  last resort” not  because  they  are  sometimes harmful to the defendant and sometimes not, depending on the circumstances of the particular case, but because they are always and unavoidably prejudicial to the  defendant.That is why decisions  like  Deck and Holbrook describe them as inherently prejudicial. Id. at 635, 125 S. Ct. at 2015; Holbrook, 475 U.S. at 568, 106 S. Ct. at 1345. Deck further recognizes that this inherent prejudice is difficult if not impossible to document. 544 U.S. at 635, 125 S. Ct. at 2015.Thus, although Strickland puts  Stephenson’s challenge  to the stun belt in a different posture, it does not change the nature of the  underlying error of requiring him to wear a visible restraint. Indeed,  Strickland itself recognizes that  nature of the  error—for example, pervasive  error versus isolated error—factors into  the prejudice analysis.466  U.S. at  695-96, 104  S. Ct. at  2069. Requiring a defendant to wear a stun belt without facts to support such a last-resort measure is a pervasive error, in that it affects the whole  trial, the jury’s perception of the defendant, and such fundamental aspects  of the  prosecution as  the presumption of innocence. The error  might not ultimately prejudice the defendant in the sense that Strickland refers to prejudice—in other  words, it might not alter the  outcome of the  trial, as when the  proof of guilt is overwhelming (see,  e.g.,  Roche, 291 F.3d at  484; Fountain v. United States, 211 F.3d 429,  436 (7th Cir.2000))—but it remains prejudicial in  the sense that  Deck and Holbrook discuss  prejudice: it undermines the presumption  of  innocence,  interferes  with  the defendant’s ability to participate  in his  own defense, and coarsens the courtroom environment. The  Strickland prejudice inquiry must begin with that recognition.