¶ 43. (dissenting). I agree with the circuit court and court of appeals and dissent on three grounds.
¶ 44. First, I agree with the court of appeals that the circuit court's determination that the search of the defendant's home was a police search should be reviewed under the clearly erroneous standard. The majority opinion's adoption of the de novo standard contravenes Wisconsin and federal law.
¶ 45. Second, whatever the standard of review, the search in this case was that of law enforcement. The search of the defendant's home without a warrant violated the Fourth Amendment.
¶ 46. Third, even if it was a probation search, the search of the defendant's home without a warrant violated the Fourth Amendment under the circumstances of this case.
¶ 47. I would affirm the court of appeals, which affirmed the circuit court's suppression order.
1 — I
¶ 48. Under existing Wisconsin case law, as well as the case law from other jurisdictions, the question of whether a probation officer acts as a stalking horse for *373law enforcement officers should be reviewed under the clearly erroneous standard.
¶ 49. First, the majority opinion's error in concluding that the question of whether a search is a police or probation search should be reviewed de novo is based on its erroneous characterization of the question before us as one of constitutional fact.
¶ 50. The majority opinion explains that a question of constitutional fact has been defined as one whose determination is decisive of constitutional rights. Majority op. at ¶ 14. I agree that this is the definition of constitutional fact.
¶ 51. Unfortunately the majority opinion does not apply this definition in the present case. The question of whether a search is a police or probation search is not decisive of constitutional rights in this case. As the court of appeals explained and as the majority opinion recognizes, the ultimate constitutional issue, that is, the constitutional fact in this case, is whether the search (whether by a probation officer or a police officer) was reasonable and conformed to the Fourth Amendment. That ultimate constitutional fact, the reasonableness of the search, should be decided de novo by this court after the relevant facts are determined by the circuit court. Whether the search is by a probation officer or police officer is not a decisive constitutional determination of reasonableness; therefore, it should be treated as an historical fact to be reviewed under the clearly erroneous, standard. As is evident in the majority opinion, the majority appears to have decided that it wishes to exercise de novo review and then labels the issue as one of constitutional fact.
¶ 52. Second, the majority opinion blithely ignores this court's most recently adopted analytical framework for determining the appropriate standard *374of review of constitutional issues. In State v. Byrge, 2000 WI 101, 237 Wis. 2d 197, 614 N.W.2d 477, this court explained that the difference between historical facts, constitutional facts, and mixed questions of fact and law is "fuzzy" at best. Thus the Byrge court decided that the standard of review of a constitutional issue should depend on the court's determination of which "judicial actor is better positioned than another to decide a matter."1 The "best institutional actor" analysis presénts a workable framework for determining the standard of review even when the issue (such as competency to stand trial) is unquestionably decisive of constitutional rights. The majority opinion ought to forthrightly adopt and apply the analysis used in Byrge. Instead, the majority opinion, without citing Byrge, much less forthrightly explaining what it is doing, applies the Byrge analysis in a back-handed way, justifying the creation of the new question of constitutional fact by noting that "independent appellate review" is important because it "provides guidance to litigants, lawyers, and trial courts." Majority op. at ¶ 18._
*375¶ 53. In Byrge the court applied the clearly erroneous standard of review to the constitutional determination of whether a defendant was competent to stand trial because "the decision pivots on factors only a trial court can appraise." Byrge, 237 Wis. 2d at ¶ 44. The circuit court's ability to observe the credibility and demeanor of the witnesses, stated the Byrge court, made the circuit court "the judicial actor best positioned to apply a legal standard to the facts." Byrge, 237 Wis. 2d at ¶ 44; see also ¶ 45.
¶ 54. I would apply the Byrge institutional analysis to the question at issue in this case. Whether the search of the defendant's home was a probation or a police search presents a fact-driven inquiry. Indeed, the particular circumstances of this case highlight the circuit court's essential role as fact-finder. Neither the probation officer nor the law enforcement officers kept any record of their communications regarding the defendant, leaving witness testimony as the only means of assessing the basis for this search.
¶ 55. The circuit court was in the best position to evaluate the testimony, weigh the witnesses' credibility, and determine whether this search was driven by law enforcement or probation objectives. By opting for de novo review, the majority opinion has rejected the circuit court's credibility assessments and weighing of the evidence.
¶ 56. Third, the majority's break from Byrge is even more surprising in light of the State's and majority opinion's failure to identify a single case in any jurisdiction that has treated the determination of whether a search is a probation or police search as anything other than a finding of fact subject to review under the clearly erroneous standard.
*376¶ 57. As the majority opinion recognizes, Wisconsin case law can be interpreted as treating this question as one of fact. Majority op. at ¶ 21.
¶ 58. Furthermore, several federal courts have applied the clearly erroneous standard of review. The majority opinion declines to follow the Ninth Circuit's standard of review, as articulated in United States v. Richardson, 849 F.2d 439 (9th Cir. 1988), and United States v. Jarrad, 754 F.2d 1451 (9th Cir. 1985). The majority contends that these cases rest on the sole issue of whether the probation officer initiated the search, whereas the majority believes a broader factual analysis is necessary. Majority op. at ¶ 23.
¶ 59. More recent Ninth Circuit cases involving probation versus police searches apply the majority opinion's broader factual analysis. Nonetheless, these cases still treat the trial court's conclusion as a question of historical fact.2 Other federal courts have *377followed the Ninth Circuit's lead.3 Several state courts have also followed the Ninth Circuit.4
*378¶ 60. The great weight of persuasive authority belies the majority's statement, majority op. at ¶ 18, that its conclusion is consistent with federal and Wisconsin precedent. Rather, it appears that the majority opinion stands alone in creating a question of constitutional fact where none existed before.
I — ! 1 — 1
¶ 61. Whether a clearly erroneous standard or a de novo standard of review is applied, the circuit court correctly ruled that the probation officer "changed hats" and was serving to advance law enforcement goals, rather than probation goals.5
¶ 62. Law enforcement officers needed probable cause and a warrant to search the defendant's home. They had neither. The search therefore violated the Fourth Amendment.
¶ 63. The majority opinion promises guidance for how probation officers, law enforcement officers, lawyers, and courts can distinguish between a probation and police search. Majority op. at ¶ 18. The guidance appears in two short conclusory paragraphs, paragraphs 32 and 34, as follows:
The five findings of historical fact clearly indicate that probation officer Hammes was cooperating *379with the law enforcement investigation of Hajicek. .. .6
[Cooperation does not change a probation search into a police search.7 . . .
[T]he search of Hajicek's residence was a probation search. . .because the probation officers conducted the search while the law enforcement officers were present at the search only for protective purposes.8
¶ 64. Does the majority opinion provide guidance about what constitutes acceptable "cooperation"? Does the majority opinion conclude that cooperation never changes a probation search into a police search? Clearly that cannot be so. I agree that cooperation does not of itself turn a probation search into a police search. Common sense tells us, however, that at some point cooperation may transform the probation officer into a stalking horse. At some point, the probation officer has, as the circuit court stated, "changed hats" and is serving a law enforcement rather than probationary function. Yet the opinion does not intimate that there are any limits on cooperation.
¶ 65. Indeed, the word "cooperation" does not appropriately describe the situation here. I agree with *380the circuit court that the probation officer "changed hats" and was serving a law enforcement rather than probationary function. The probation officer allowed law enforcement to dictate his operations. The probation officer agreed to allow law enforcement officers to interfere with ordinary probationary supervision and to delay a probationary search for more than three weeks. Law enforcement officers were not able to obtain a search warrant because their information was not good enough and then the probation officer worked with these same officers on a "probation" search looking for the same evidence as the aborted search warrant would have sought.9 •
¶ 66. Does the majority opinion provide guidance about how to distinguish a probation search from a law enforcement search? Does the majority opinion conclude that so long as the probation officers, not the police officers, physically conduct the search with the law enforcement officers present for protective purposes, the search is a probation search? Clearly that cannot be so. Such a rule would put form over substance and make a mockery of the stalking horse doctrine. But that's what the opinion appears to say.
¶ 67. Guidance? I think not. Puzzlement? I think so. The question the readers should ask, after reading the majority opinion, is whether they can hypothesize any realistic fact situation in which a probation officer who performs a search under the protection of the very law enforcement officers with whom they have been "cooperating" would be transformed into a stalking horse. I fear that no such situation exists.
¶ 68. When the probation officer is serving law enforcement objectives, as is the case here, a search
*381warrant is necessary under the Fourth Amendment. I would therefore affirm the circuit court's suppression order.
I — I i — I j.
¶ 69. Even viewing the search as a probation search, I conclude that the warrantless search of the defendant's home was unreasonable and violated the Fourth Amendment. The State bears the burden of persuasion in this case on the reasonableness of the search without a warrant.
¶ 70. The majority opinion's conclusion that this probation search was reasonable relies on the U.S. Supreme Court's decision in Griffin v. Wisconsin, 483 U.S. 868 (1987), as well as the agency regulations regarding probation searches. Reliance on both of these authorities is misplaced.
¶ 71. In Griffin, the Supreme Court articulated two reasons for declining to require probation officers to obtain a search warrant. Neither of these reasons applies to the facts of this case.
¶ 72. First, the Griffin Court stated that "[a] warrant requirement would interfere to an appreciable degree with the probation system, setting up a magistrate rather than the probation officer as the judge of how close a supervision the probationer requires." Griffin, 483 U.S. at 876. Here, however, law enforcement officers interfered with the probation officer's plan to search the defendant's home immediately.10 Moreover, the probation officer opted to forgo all supervisory activities in order to avoid alerting the defendant to the *382law enforcement investigation. The need to avoid interference with probationary objectives was not present in this case because law enforcement had already interfered. Therefore Griffin s narrow exception to the warrant requirement does not apply.
¶ 73. Second, the Griffin Court stated that "the delay inherent in seeking a warrant would make it more difficult for probation officers to respond quickly to evidence of misconduct. . .and would reduce the deterrent effect that the possibility of expeditious searches would otherwise create. . .." Griffin, 483 U.S. at 876 (citations omitted). Yet here, the probation officer allowed law enforcement to delay his intended search for more than three weeks, three weeks during which the probation officer deliberately avoided taking any action whatsoever to supervise the defendant in a manner consistent with his probationary objectives. The need to avoid delay in responding to evidence of a probationer's misconduct was not present in this case. Griffin s narrow exception to the warrant requirement does not apply.
¶ 74. The majority does not explain how the probation officer's failure to take any supervisory action whatsoever for almost a month is consistent with probationary objectives, objectives which the Griffin Court described as "protecting the public interest" and having in mind "the welfare of the probationer (who in the regulations is called a client)." Griffin, 483 U.S. at 876. When the probation officer's actions are fundamentally inconsistent with the important probationary objective of active supervision, Griffin's narrow exception allowing warrantless searches that advance probationary objectives no longer applies.
¶ 75. The majority opinion also relies on the probation officer's compliance with the Wisconsin *383Administrative Code. This reliance is misplaced as well. The majority contends that the probation officer followed Wis. Admin. Code § 328.21(7)(b) in searching Hajicek's home. Majority op. at ¶ 40. Further, the majority has found that the more than three-week delay in searching the defendant's home did not violate any provision of the Wisconsin Administrative Code. Majority op. at ¶ 41.
¶ 76. The majority opinion overlooks the fact that the probation officer violated Wis. Admin. Code § 328.30, which requires probation officers to document all of their communications regarding a case, including "maintain[ing] a chronological log of all case related contacts." Here, the defendant's probation officer documented none of his communications with law enforcement officials. When asked during the suppression hearing why he failed to document any of his communications, in violation of Wis. Admin. Code § 328.30, the probation officer answered simply, "I just didn't." The majority opinion's conclusion also ignores the uncontradicted testimony of Professor Walter Dickey, the former head of the Department of Corrections and the drafter of the precursor of the applicable Wisconsin Administrative Code rules, regarding the violations that occurred in this case.
¶ 77. The probation officer's violations of administrative procedures further undermine the majority's conclusion that the officer was acting reasonably and in conformance with legitimate probation objectives when he searched the defendant's home without a warrant.
¶ 78. For the reasons set forth, I dissent.
*384¶ 79. I am authorized to state that Justices WILLIAM A. BABLITCH and ANN WALSH BRADLEY join this dissent.
State v. Byrge, 2000 WI 101, ¶ 39, 237 Wis. 2d 197, 614 N.W.2d 477. See also State v. Garfoot, 207 Wis. 2d 214, 234, 558 N.W.2d 626 (1997) (C.J. Abrahamson concurring) ("The standard for appellate review of an issue thus depends on a determination of whether an appellate court or a trial court is the more appropriate and competent forum to make the particular decision."); State v. McMorris, 213 Wis. 2d 156, 165-66, 570 N.W.2d 384 (1997) (treating the issue of independent source for in-court identification like the issue of purging evidence of the taint of illegal search as a constitutional fact; allowing independent review to enable appellate courts to flesh out the rule and provide guidance to litigants, lawyers, and trial and appellate courts and to provide uniformity in constitutional decision-making).
For example, in United States v. Watts, 67 F.3d 790 (9th Cir. 1995), rev'd on other grounds, 519 U.S. 148 (1997), the court of appeals stated that ”[t]he appropriate inquiry is whether the probation officer used the probation search to help police evade the Fourth Amendment's usual warrant and probable cause requirements or whether the probation officer enlisted the police to assist his own legitimate objectives." Even with the broader fact inquiry, the standard of review remained clearly erroneous. Watts, 67 F.3d at 794.
At least two other circuits have followed the Ninth Circuit in treating the question of probation versus police search as a question of fact reviewable under the clearly erroneous standard. See, e.g., United States v. Cardona, 903 F.2d 60, 65 (1st Cir. 1990) ("The law will not allow a parole officer to serve as a cat's paw for the police. . . .When and if the integrity of a challenged action is controverted, the dispute is determinable as a question of fact on a case-by-case basis."); United States v. McFarland, 116 F.3d 316, 318 (8th Cir. 1997) ("Parole and police officers may work together, however, provided the parole officer is pursuing parole-related objectives and is not merely a 'stalking horse' for the police. . . .In this case, the district court found [the probation officer] authorized the police to carry out the challenged searches to determine if McFarland was violating his parole. This factual finding is not clearly erroneous.") (citations omitted).
See, e.g., State of Hawaii v. Proprios, 879 P.2d 1057, 1064 (Haw. 1994) ("Notwithstanding the existence of an objectively valid probationary purpose, we hold that a warrantless probationary search is unreasonable if it is conducted for a subjectively improper purpose. As indicated previously, whether a particular search is improper 'is a question of fact subject to the clearly erroneous standard of review.'"); State v. Cowans, 717 N.E. 2d 298, 307 (Ohio 1999) ("Based upon the testimony presented and the trial court's assessment of the credibility of the witnesses, the trial court made a factual finding that [the probation officer] was not acting as a stalking horse for the deputies. Instead, the court found, she 'had her own objectives in conducting the search.' We are bound by that finding unless the record contains insufficient evidence to support it."); Commonwealth v. Williams, 692 A.2d 1031, 1037 n.11 (Pa. 1997) ("[E]vidence should be suppressed if the parole agent switches hats and acts as a 'stalking horse' for the police by circumventing the requirement for a warrant. . . .Here, the suppression court explicitly found that the parole agent who *378conducted the search was not acting in such a capacity.. . .[N]o evidence in the record even remotely suggests that the trial court erred in reaching this conclusion.").
The circuit court correctly concluded that "[allthough there may be facts and exigent circumstances in a different case which would justify a probation officer's decision not to supervise a client at the behest of law enforcement, this is not such a case."
Majority op. at ¶32. The circuit court's findings upon which the majority opinion bases its legal determination that a probation search occurred are as follows: Law enforcement officers dictated the timing of the search, participated in the search, and notified the assistant district attorney before the search; the police officers and probation officer failed to document their communications with each other; the probation officer did not carry out the objectives of probation supervision and failed to supervise the probationer for more than three weeks during the delay of the search. Majority op. at ¶ 28.
Majority op. at ¶ 32.
Majority op. at ¶ 34.
In most instances, in contrast to this case, law enforcement asks the probation officer to conduct a search.
See circuit court order at 7 ("Although Hammes felt he had good cause to do a probation search, he did not do so at the behest of law enforcement.").