FILED
United States Court of Appeals PUBLISH Tenth Circuit
UNITED STATES COURT OF APPEALS June 2, 2021
Christopher M. Wolpert
FOR THE TENTH CIRCUIT Clerk of Court
_________________________________
UNITED STATES OF AMERICA, Plaintiff - Appellee, v. No. 19-1487 PERRY WAYNE SUGGS, JR.,
Defendant - Appellant.
_________________________________
Appeal from the United States District Court for the District of Colorado (D.C. No. 1:18-CR-00089-WJM-1)
_________________________________
Daniel T. Hansmeier, Appellate Chief (Melody Brannon, Federal Public Defender, with him on the briefs), Kansas City, Kansas, for Defendant–Appellant.
Karl L. Shock, Assistant United States Attorney (Jason R. Dunn, United States Attorney, with him on the brief), Denver, Colorado, for Plaintiff–Appellee.
_________________________________
Before MATHESON, BALDOCK, and MORITZ, Circuit Judges.
_________________________________
BALDOCK, Circuit Judge.
_________________________________
This appeal stems from an incident at the corner of a crosswalk in Colorado Springs. It is an unusual product of a familiar conflict: the pedestrian and the right- hand turn. A pedestrian wanted to cross the street. At the same time, the driver of a vehicle wanted to turn right. Words were exchanged. Then the driver pulled out a
2
gun and took a shot at the pedestrian. Fortunately the bullet didn’t strike anyone. The vehicle sped off, and the pedestrian called 911. Law enforcement focused their investigation on Defendant Perry Suggs. Warrants were issued, Defendant’s home was searched, and incriminating evidence was discovered.
A federal grand jury charged Defendant with possession of a firearm and ammunition by a convicted felon, in violation of 18 U.S.C. § 922(g)(1). Before trial, Defendant moved to suppress the evidence found during sequential searches of his home and an SUV parked under his carport. He argued that the warrant to search his home violated the Fourth Amendment’s particularity requirement and that officers would not have found the evidence used against him but for the invalid warrant. The district court disagreed and denied the motion. Defendant now appeals that decision.
We have jurisdiction under 28 U.S.C. § 1291 to consider whether the district court erred when it denied Defendant’s motion to suppress. We hold that it did. Because the residential search warrant failed to meet the Fourth Amendment’s particularity requirement and cannot be saved by the severability doctrine, we vacate the order denying Defendant’s suppression motion. Yet the question remains whether the good-faith exception to the exclusionary rule saves the incriminating evidence from suppression. We remand for the district court to resolve underlying factual disputes and consider the remedial question in light of this opinion.
I.
What started out as a verbal altercation at the corner of a crosswalk in Colorado Springs quickly escalated when the driver of a vehicle pulled out a gun and
3
fired a shot at a pedestrian’s feet. Officer Adam Menter responded to the incident. He interviewed witnesses and identified Defendant as the driver and owner of the vehicle involved in the shooting.
After concluding his initial investigation, Officer Menter got an arrest warrant for Defendant. He also applied for a search warrant with the state district court. As part of the application, Officer Menter submitted an affidavit (labeled as Attachment A) that detailed the circumstances of the vehicle shooting and the fruits of his investigation. In addition, the affidavit noted that Defendant was a confirmed gang member, suspected to be involved in several other crimes, and had prior felony convictions for menacing and possession of a weapon by a previous offender. The warrant application requested authority to search Defendant’s home for the items listed in “Attachment B,” which described the targeted property as follows:
4
Suppl. ROA, Vol. I at 52.
The state district court issued a warrant that identified the place to be searched as Defendant’s home. As for the items to be searched for and seized, the warrant incorporated by reference the same Attachment B that had accompanied the warrant application. Officer Menter’s affidavit (Attachment A), on the other hand, was not expressly incorporated into the warrant.
While conducting surveillance in anticipation of the search, officers observed Defendant drive away in the vehicle involved in the shooting. After Defendant parked the vehicle at a nearby gas station, members of the SWAT team arrested him. Officers then went to Defendant’s home to execute the search warrant.
Led by Officer Teresa Tomczyk, the SWAT team conducted a protective sweep of Defendant’s home and then cleared the outside surrounding area of any threats. During this process, Officer Tomczyk shined her flashlight through the window of an SUV (not the vehicle involved in the shooting) parked under the carport, looking for persons who might be hiding. She did not see any people. What she saw instead were two guns, a magazine, and two handgun cases.
With the premises secure, officers conducted the search of Defendant’s home.
One officer found and seized a box of ammunition that matched the ammunition used in the vehicle shooting. Officer Menter discovered a bank statement linking Defendant to the residence. He also found and seized a retail sale contract for the vehicle involved in the shooting.
5
At some point, Officer Tomczyk told Officer Menter about the guns she saw in the SUV parked under the carport. After taking a look for himself, Officer Menter returned to the police station and used this information to obtain a warrant to search the SUV. This warrant was almost identical to the residential search warrant. Officer Menter then returned to Defendant’s home and executed the vehicle warrant. During the search of the SUV, Officer Menter found and seized a handgun, a rifle, ammunition, gun carrying cases, and the vehicle’s registration document.
A grand jury indicted Defendant on one count of being a felon in possession of a firearm and ammunition, in violation of 18 U.S.C. § 922(g)(1). Defendant filed a motion to suppress the evidence found in his home and in the SUV. He argued that the residential search warrant was invalid because it violated the Fourth Amendment’s particularity requirement and that the evidence found in the SUV should be suppressed as fruit of the initial, unconstitutional search. The district court denied Defendant’s motion.
After trial, the jury convicted Defendant. The district court sentenced him to 90 months’ imprisonment, to be followed by a 3-year term of supervised release. Defendant now appeals the denial of his suppression motion.
II.
We accept the district court’s factual findings unless they are clearly erroneous, but we review de novo the district court’s ultimate determination of reasonableness under the Fourth Amendment. United States v. Burgess, 576 F.3d 1078, 1087 (10th Cir. 2009).
6
III.
As he did before the district court, Defendant contends that the residential search warrant lacked particularity, rendering the search of his home unconstitutional and requiring suppression of the evidence discovered as a result of the initial illegality. Meanwhile, the Government argues that the residential search warrant was valid and that, even if it wasn’t, both the severability doctrine and the good-faith exception to the exclusionary rule save the evidence from suppression.
We begin our analysis by considering the validity of the residential search warrant and determine that it violated the Fourth Amendment’s particularity requirement. We next consider whether the severability doctrine applies and conclude that the warrant is not a good candidate for this remedy. Finally, rather than decide whether officers relied on the invalid warrant in good faith, we remand for the district court to resolve underlying factual disputes and fully address the issue in the first instance.
A.
Defendant first argues that the search warrant for his home was invalid because it failed to describe with particularity the things to be searched for and seized. We agree.
1.
Our analysis begins, as it must, with the Fourth Amendment’s text. The Fourth Amendment provides that “no Warrants shall issue” without “particularly describing the place to be searched, and the persons or things to be seized.” U.S.
7
Const. amend. IV. This particularity requirement, as it has come to be known, was most immediately the product of contemporary revulsion against a regime of general warrants that gave British officials carte blanche to search and seize property of American colonists. Stanford v. Texas, 379 U.S. 476, 481–82 (1965). The Framers constitutionalized the requirement to ensure that a “search will be carefully tailored to its justifications” and “will not take on the character of the wide-ranging exploratory searches” English kings once favored. Maryland v. Garrison, 480 U.S. 79, 84 (1987). To this end, a search warrant must “describe the items to be seized with as much specificity as the government’s knowledge and circumstances allow.” United States v. Leary, 846 F.2d 592, 600 (10th Cir. 1988).
The warrant here targeted some particular items but also included a catch-all phrase authorizing the search and seizure of “[a]ny item identified as being involved in crime.” Wisely, the Government does not contest that this language, on its face, would be anything but a license for the sort of general rummaging outlawed by the Fourth Amendment. The Government’s claim, instead, is that under a natural reading of the warrant, “[a]ny item identified as being involved in crime” actually means “any item identified as being involved in the vehicle shooting under investigation.”
It is true, as the Government emphasizes, that we construe search warrants in a practical and commonsense fashion, avoiding a hypertechnical reading of their terms. Leary, 846 F.2d at 600. A search warrant need not meet the rigors of Strunk, White, Roget, Merriam, and Webster to satisfy the particularity requirement. United States v. Otero, 563 F.3d 1127, 1132 (10th Cir. 2009). But the problem here is that no
8
reasonable construction of the residential search warrant—be it technical, practical, or otherwise—can sustain the Government’s interpretation.
For one thing, the word “crime” does not necessarily mean a single act or specific incident constituting an offense punishable by law, as the Government suggests. Depending on how the word is used, “crime” means such acts “collectively,” Webster’s New World College Dictionary 328 (3d ed. 1997), or “criminal activity” in general, Random House Dictionary of the English Language 476 (2d unabridged ed. 1987); see also American Heritage Dictionary of the English Language 313 (New College ed. 1978) (defining “crime” as “[u]nlawful activity in general”). One might say, for example, that “[t]he profits of crime are untaxed,” Crime, Oxford English Dictionary (3d ed. 2010) (accessed online) (defining “crime” as criminal “acts collectively”), or that “efforts to fight crime” too often go unrewarded, Crime, Merriam-Webster.com Dictionary, https://www.merriam- webster.com/dictionary/crime (last visited April 27, 2021) (defining “crime” as “criminal activity”); see also Random House Dictionary of the English Language, supra, at 476 (noting the same usage). In those sentences, the word “crime” does not mean a specific act or incident punishable by law, but criminal activity in general. So too with the search warrant for Defendant’s home.
The residential search warrant’s catch-all phrase does not refer to “the crime,”
which could indicate a nexus exists between “[a]ny item” sought and the specific offense that prompted the warrant. Nor does the catch-all phrase contain the words “a crime,” which might suggest, given proper context, that it covers all suspected
9
offenses specified elsewhere. The residential search warrant, instead, uses the prepositional phrase “in crime,” which more naturally means criminal activity in general. See Pennsylvania v. Ashe, 302 U.S. 51, 54–55 (1937) (using the phrase “in crime” to mean crime generally). Thus, the warrant’s plain language authorized officers to search for and seize evidence of any crime.
There is more. The language and structure of the search warrant, taken as a whole, indicate that the phrase “[a]ny item identified as being involved in crime” has broad meaning. Those words are not tacked onto or linked with any of the other operative sentences describing the items to be searched for and seized. Rather, they sit by themselves under Attachment B’s “MISCELLANEOUS” heading, which suggests they authorized officers to search for items different than those expressly listed elsewhere. “Differences such as subject headings and paragraph formation might seem insignificant, but if we are to follow our command of reading each part of the warrant in context, these structural indicators are useful tools.” Otero, 563 F.3d at 1133. And when these structural features are considered in the context of the overall warrant, they indicate the catch-all provision was subject to no affirmative limitations other than that the items sought must be evidence of criminal activity.
A notable omission in the language of the residential search warrant confirms this interpretation. When a warrant does not otherwise describe the evidence to be searched for and seized with sufficient particularity, that gap can sometimes be filled if the warrant specifies the crime under investigation. See, e.g., Burgess, 576 F.3d at 1083 (upholding a warrant authorizing a search for “evidence to show the
10
transportation and delivery of controlled substances”); United States v. Brooks, 427 F.3d 1246, 1252 (10th Cir. 2005) (upholding a warrant authorizing officers to search for “evidence of child pornography”). In other words, a warrant may satisfy the particularity requirement if its text constrains the search to evidence of a specific crime such that it sufficiently narrows language that, on its face, sweeps too broadly. Compare United States v. Robertson, 21 F.3d 1030, 1032 (10th Cir. 1994) (concluding that a warrant authorizing agents to search for and seize four items specifically mentioned “and other instrumentalities and fruits of the crime of armed carjacking” was sufficiently particularized), with Leary, 846 F.2d at 601–02 (holding that a reference to two federal statutes did not sufficiently limit the search because those statutes cover a broad range of activity).
Had the warrant here specified that the search of Defendant’s home was being undertaken in connection with the vehicle shooting, it might be possible to read the catch-all phrase, in context, as authorizing a search only for evidence related to that crime. But the warrant didn’t do that. Nowhere does the warrant reference any specific offense—let alone the particular firearm-related crime under investigation. In fact, the warrant never even mentions the vehicle shooting. To adopt the Government’s interpretation, then, we would have to rewrite the warrant to include what it lacks: a reference to the targeted crime(s). We of course can’t do that.
The Government tries to head off this conclusion by arguing that its construction of the residential search warrant is consistent with the Supreme Court’s decision in Andresen v. Maryland, 427 U.S. 463 (1976). But the Government’s
11
reliance on Andresen is misplaced. In Andresen, the challenged warrants listed several items related to a specific real estate transaction involving “Lot 13T” and then closed with the phrase: “together with other fruits, instrumentalities and evidence of crime at this (time) unknown.” Id. at 479. Immediately preceding the allegedly problematic catch-all phrase, the warrant referred to the specific crime under investigation: “the crime of false pretenses, in violation of Article 27, Section 140, of the Annotated Code of Maryland.” Id. at 480 n.10. Accordingly, the Court found “it clear from the context that the term ‘crime’ in the warrants refers only to the crime of false pretenses with respect to the sale of Lot 13T.” Id. at 480–81.
Andresen lends no succor to the Government’s position here because the search warrant for Defendant’s home does not tie any of the items sought for seizure to a specific crime. All the warrant says is that probable cause existed to believe the things to be searched for and seized either “ha[d] been used as a means of committing a criminal offense,”
were “illegal to possess,”
“[w]ould be material evidence in a subsequent criminal prosecution, or required, authorized or permitted by a statute of the State of Colorado,”
“[were] kept, stored, transported, sold, dispensed, or possessed in violation of state law under circumstances involving a serious threat to the public safety, or order, or to the public health, or”
“would aid in the detection of the whereabouts of or in the apprehension of a person for whom a lawful arrest order is outstanding.”
12
Suppl. ROA, Vol. I at 47 (emphasis added). This broad, disjunctive language provides no context from which to constrain the search to evidence of a specific crime. See Cassady v. Goering, 567 F.3d 628, 635 (10th Cir. 2009).
Read as a whole, the warrant told officers they could search for evidence of any crime rather than only evidence related to the vehicle shooting. So open-ended is the description that the warrant can only be described as a general one, akin to the instruments of oppression vivid in the memory of newly independent Americans when the Fourth Amendment was adopted. See Garrison, 480 U.S. at 84. We addressed a search warrant with a similar catch-all phrase in Cassady v. Goering. There, we invalidated a warrant that listed “specific items mostly related to a narcotics operation” but also contained a catch-all phrase authorizing officers to search for “all other evidence of criminal activity.” 567 F.3d at 635. We have no basis to reach a different conclusion here. For the reasons just given, the warrant to search Defendant’s home flubbed the Fourth Amendment’s particularity requirement.
2.
The Government urges us to reach a contrary conclusion about the validity of the residential search warrant. It gives several reasons why the affidavit submitted in support of the warrant application, which details the vehicle shooting and the investigation that followed, supplies the required particularity. But not one of them holds water.
A supporting affidavit can sometimes cure a warrant’s lack of particularity.
Leary, 846 F.2d at 603. But to remedy the defect, two requirements must be met:
13
(1) the warrant and the affidavit must be attached; and (2) the warrant must expressly incorporate the affidavit. Id. As to the latter requirement, a warrant does not incorporate an affidavit merely by mentioning the affidavit or reciting that the magistrate judge found probable cause to authorize the search. See Groh v. Ramirez, 540 U.S. 551, 557–58 (2004). Instead, “the search warrant must expressly refer to the affidavit and incorporate it by reference using suitable words of reference.” United States v. Williamson, 1 F.3d 1134, 1136 n.1 (10th Cir. 1993) (emphasis added) (quoting Leary, 846 F.2d at 603).
Officer Menter’s affidavit may have been physically attached to the search warrant, as the district court found, but the warrant did not incorporate the affidavit. The warrant mentions the affidavit only once, noting that Officer Menter “has made an Application and Affidavit to the Court for the issuance of a Search Warrant.” To be sure, this lone notation does not suggest the affidavit was incorporated into the warrant. And nowhere else does the warrant use suitable words of reference to expressly incorporate the affidavit.
The Government conceded as much before the district court. But now it reverses course, claiming the warrant incorporated the affidavit because it provides that “probable cause is found for the issuance of a Search Warrant to search the person(s) or premises specified in the application.” That language, the Government insists, incorporated the supporting affidavit under this court’s “intervening precedent” in United States v. Sadlowski, 948 F.3d 1200 (10th Cir. 2020). Appellee’s Br. at 26. We are not persuaded.
14
Even if Sadlowski qualifies as an “intervening” precedent in this context—a dubious proposition, see United States v. Titties, 852 F.3d 1257, 1264 n.5 (10th Cir. 2017)—it does not help the Government here. The warrant in Sadlowski not only authorized officers to search “the persons and/or place described in the affidavit,” 948 F.3d at 1205 (emphasis added) (brackets omitted), but also provided: “A copy of the Affidavit is attached and made a part of this Search Warrant,”1 Mot. to Suppress Evid., Ex. 1, United States v. Sadlowski, No. 1:16-cr-00847-JCH (D.N.M. Aug. 18, 2016), ECF No. 38-1. Given such language, it is unsurprising we said that “the warrant plainly incorporated the underlying affidavit.” Sadlowski, 948 F.3d at 1205. In visible contrast to Sadlowski, the warrant to search Defendant’s home plainly did not incorporate the supporting affidavit.
The rest of the language in the residential search warrant confirms this conclusion. Although the warrant did not expressly incorporate Attachment A (the affidavit) by reference, it did the opposite with Attachment B. As for the things to be searched for and seized, the warrant provides: “See Attachment ‘B’ which is hereby incorporated in reference.” This language, as a matter of common sense as well as logic, shows that the warrant’s drafter knew how to use appropriate words of incorporation. Cf. United States v. Sanders, 796 F.3d 1241, 1250 (10th Cir. 2015) (invoking the negative-implication canon, under which the expression or inclusion of
1 Although the search warrant in Sadlowski is not part of the record on appeal in this case, we take judicial notice of it under Federal Rule of Evidence 201. United States v. Duong, 848 F.3d 928, 930 n.3 (10th Cir. 2017) (“taking judicial notice of all filings of the parties and rulings of the district court” in a different case).
15
one thing implies the exclusion of others, in the Fourth Amendment context). It also dispels any doubt that the affidavit was not expressly incorporated into the warrant.2 Relying on our decision in United States v. Ortega-Jimenez, 232 F.3d 1325 (10th Cir. 2000), the Government says the unincorporated affidavit can still cure the facially invalid warrant to search Defendant’s home. This is so, the Government maintains, because the officer who prepared the affidavit was among the officers who executed the search. There are several problems with this argument.
First, and foremost, the Fourth Amendment by its terms requires that the warrant—not the supporting documents presented to the judicial officer asked to issue the warrant—particularly describe the things to be seized. U.S. Const. amend. IV. And for good reason: the description of what is sought to be seized is written in, or incorporated into, the warrant to ensure that the executing officer(s) stay within the specific bounds set by the issuer. Leary, 846 F.2d at 600. That important function would not necessarily be vindicated if we allowed an unincorporated affidavit, rather than the warrant itself, to dictate what can be searched for and seized. To the contrary, doing so would create a danger that officers will conduct a search that goes beyond constitutional limits. By the same token,
2 In addition to signing the residential search warrant itself, the issuing judge signed Officer Menter’s affidavit. The state judge’s signature at the bottom of the affidavit cannot substitute for express incorporation or otherwise save the warrant from facial invalidity. See United States v. Russian, 848 F.3d 1239, 1243, 1246–47 (10th Cir. 2017) (concluding that a warrant lacked particularity even though it referenced the supporting affidavit, which was signed by the magistrate judge and particularly described the place to be searched and the things to be seized).
16
without clear incorporation of an affidavit into a search warrant, we cannot simply assume the former shapes the latter. Otherwise, we might indeed elevate the author of the unincorporated affidavit over the warrant’s issuer, thereby transforming the independent magistrate into little more than a rubber stamp.
Second, Ortega-Jimenez has little, if any, bearing on the question before us because it involved a dispute about whether a search exceeded the scope of a “facially valid” warrant’s particularized terms, not whether the warrant lacked particularity. 232 F.3d at 1329. The Government attempts to brush aside this distinction, suggesting that scope and particularity are two sides of the same coin. Not so. See United States v. Angelos, 433 F.3d 738, 745–46 (10th Cir. 2006) (treating particularity and scope as separate inquiries and concluding that a search conducted pursuant to a warrant that “met the Fourth Amendment’s particularity requirement” was unconstitutional because it exceeded the warrant’s scope). We have never applied Ortega-Jimenez in the particularity context, and we decline the Government’s invitation to do so today.
Third, Ortega-Jimenez’s suggestion that “an officer’s knowledge of the terms of [an] affidavit [can] cure a lack of particularity on the face of a warrant,” 232 F.3d at 1329, is based on a misreading of our decision in United States v. Simpson, 152 F.3d 1241, 1248 (10th Cir. 1998). Simpson addressed whether a search went beyond the scope of a warrant that contained an internal inconsistency about the place to be searched, which was described in detail, not whether the warrant lacked particularity. 152 F.3d at 1247–48. Although Simpson noted, in dicta, that the affidavit supported
17
the district court’s conclusion that the officer acted “in good faith,” we did not suggest that an unincorporated affidavit can save an insufficiently particularized warrant. See id. at 1248. On the contrary, this court declared just the opposite: that an affidavit “not physically attached to the warrant nor specifically incorporated into the warrant . . . cannot cure a lack of particularity.” Id. (citing United States v. Dahlman, 13 F.3d 1391, 1395 (10th Cir. 1993)).
Finally, to the extent Ortega-Jimenez can be read to hold otherwise, it is at odds not only with Simpson itself but with earlier, settled precedent. In Leary, for example, the law enforcement agent who produced the supporting affidavit also executed the search warrant. 846 F.2d at 594. Yet we refused to consider the affidavit when determining whether the warrant lacked particularity. Id. at 603–04. So too in Williamson. 1 F.3d at 1135, 1136 & n.1. “In cases of conflicting circuit precedent our court follows earlier, settled precedent over a subsequent deviation therefrom.” United States v. Sabillon-Umana, 772 F.3d 1328, 1334 n.1 (10th Cir. 2014) (cleaned up). And as earlier, settled precedent makes clear here, particularity in Officer Menter’s affidavit could not cure the defect in the warrant for Defendant’s home because the warrant did not expressly incorporate the affidavit.
B.
Although the residential search warrant’s catch-all phrase sweeps too broadly, this constitutional infirmity does not necessarily doom the entire warrant. Sometimes we can apply the doctrine of severability to save a facially invalid warrant. United States v. Sells, 463 F.3d 1148, 1154–55 (10th Cir. 2006). Under this doctrine, we
18
sever the offending portion from the warrant, suppress any evidence collected under it, and admit the evidence collected under the valid portions that remain. Id. at 1155.
We apply a multistep analysis to determine whether a warrant is severable.
First, we divide the warrant in a commonsense way. Id. at 1155–56. Then we examine the validity of each section. Id. at 1156–57. If at least one section passes constitutional muster—meaning it satisfies the probable cause and particularity requirements of the Fourth Amendment—we determine whether the valid parts are distinguishable from the invalid ones. Id. at 1158. When “each of the categories of items to be seized describes distinct subject matter in language not linked to language of other categories, and each valid category retains its significance when isolated from [the] rest of the warrant, then the valid portions may be severed from the warrant.” Id.
“But just because the portions of a warrant can be severed does not necessarily mean we may sever them in all instances.” United States v. Cotto, 995 F.3d 786, 799 (10th Cir. 2021). Even when a part of a warrant is valid and distinguishable, blanket suppression may still be required if “the invalid portions so predominate the warrant that the warrant in essence authorizes a general, exploratory rummaging in a person’s belongings.” Cassady, 567 F.3d at 638 (quoting Sells, 463 F.3d at 1158). In those situations, courts may admit seized items only when the valid and distinguishable portions “make up the greater part of the warrant.” Sells, 463 F.3d at 1158 (quoting United States v. Naugle, 997 F.2d 819, 822 (10th Cir. 1993)).
19
This “greater part” inquiry “focuses on the warrant itself rather than upon an analysis of the items actually seized during the search.” Id. at 1159. In conducting the inquiry, we use a commonsensical approach that examines both the quantitative and qualitative aspects of the valid parts of the warrant relative to the invalid parts. Id. at 1160. So while we will tally the valid and invalid parts of the warrant, we must also examine “the relative scope and invasiveness” of each part. Id.
Applying the severability framework here, the residential search warrant already did the first step for us. The warrant is divided into four separate categories of items to be searched for and seized: “GENERAL INFO”; “GUNS INVOLVED”; “VEHICLE”; and “MISCELLANEOUS.” Because this division is both logical and sensible, we have no reason to divide the warrant further. See Naugle, 997 F.2d at 820–22 (dividing a list of items into four general categories); Cassady, 567 F.3d at 638 (dividing the warrant “into three general parts”). Breaking down Attachment B’s sections so that each of their thirteen bullet points is a separate part, as the Government suggests, would only risk an improper “hypertechnical” reading of the warrant. See Sells, 463 F.3d at 1156.
Turning to step two, we evaluate each section of the warrant to determine whether it is valid. Because there is no dispute the “general info” section is valid, we will assume that it is. And we already concluded that the “miscellaneous” section, which authorized officers to search for and seize “[a]ny item identified as being involved in crime,” is invalid because it lacks particularity. That leaves the “guns involved” and “vehicle” sections.
20
We agree with Defendant that the “vehicle” section is invalid. This section authorized officers to search for and seize “[i]ndicia of ownership of vehicle” and “[v]ehicle registration.” Use of the singular “vehicle” suggests that this part of the warrant targeted only evidence related to one vehicle. But it does not identify any particular vehicle, even though Officer Menter had specific information about the vehicle involved in the shooting when he authored the warrant application. As a result, the warrant did not “describe the items to be seized with as much specificity as the government’s knowledge and circumstances allow.” Leary, 846 F.2d at 600.
We disagree with Defendant about the “guns involved” section of the warrant.
This section authorized officers to search for “[a]ny and all firearms” and “any and all ammo,” as well as five other types of items related to possessing a firearm. In United States v. Pulliam, we held that a warrant authorizing officers to search for, among other things, “[a]ny and all firearms and ammunition” was sufficiently particularized because the defendant was a felon and felons cannot legally possess firearms. 748 F.3d 967, 972 (10th Cir. 2014). A similar conclusion fits this case.
Both Defendant’s suspected involvement in the vehicle shooting and status as a convicted felon were disclosed in the affidavit submitted as part of the warrant application. The affidavit also noted that Defendant “is a confirmed gang member with the Gangster Disciples,” has prior felony convictions involving weapons, and was the suspect in several other investigations, including two felony assaults. Officer Menter concluded his affidavit by stating that, based on his training and experience, “parties with gang affiliations and a history of violent criminal behavior” are “likely”
21
to “possess more than one firearm.” Given this information, the “guns involved” section particularly describes evidence for which probable cause existed: any and all firearms and ammunition, along with related paraphernalia. “No specific description of a gun [or ammo] was necessary.” Pulliam, 748 F.3d at 972.
Our next step is to distinguish the valid parts of the warrant from the invalid parts. That’s an easy task when, like here, the categories of items targeted by the warrant make no reference to any of the other categories and all deal with different types of property. Cotto, 995 F.3d at 800. The first and second categories of items (“general information” and “guns involved”) to be searched for and seized under the warrant are valid and distinguishable from the third and fourth categories of the warrant (“vehicle” and “miscellaneous”) that are invalid.
The only remaining question as to severability is whether the valid portions “make up the greater part of the warrant.” Sells, 463 F.3d at 1160. They do not. Because the warrant has two valid and two invalid sections, we might have had a draw if the “greater part” inquiry was limited to merely counting parts. But given the scope and invasiveness of the invalid parts—particularly the “miscellaneous” catch- all section authorizing officers to search for and seize “[a]ny item identified as being involved in crime”—the valid parts do not constitute the greater part of the warrant.
Cassady, once again, is instructive. There we divided the warrant into “three general parts: (1) the section authorizing seizure of narcotics and related illegal contraband; (2) the section authorizing seizure of all other evidence of criminal activity; and (3) the section authorizing seizure of [the defendant’s] personal property
22
if its seizure is authorized on a number of enumerated grounds totally unrelated to a narcotics operation.” 567 F.3d at 638. We assumed the first section was valid and found the second and third sections invalid. Id. at 638–39. We then held that “the invalid portions of the warrant [were] sufficiently ‘broad and invasive’ so as to ‘contaminate the whole warrant.’” Id. at 641 (quoting Sells, 463 F.3d at 1151). That’s because the invalid portions “permitted search and seizure of any evidence of any crime,” id. at 642, “whether or not related to [the underlying crime], and largely subsume[d] those provisions that would have been adequate standing alone,” id. at 641 (quoting Voss v. Bergsgaard, 774 F.2d 402, 406 (10th Cir. 1985)).
The same circumstances are present here. The residential search warrant’s catch-all section authorized the search and seizure of evidence of any crime—“be it murder, robbery, stolen property, fraud, tax evasion, or child pornography, to name just a few examples.” See id. at 642. Thus, the invalid portions of the search warrant are just as broad and invasive as the tainted provisions in the Cassady warrant. So even though some portions of the warrant to search Defendant’s home are valid and distinguishable, those portions do not make up the greater part of the warrant.
The Government makes two contrary arguments, both unconvincing. First, it attempts to distinguish Cassady as a civil suit where the social cost of total suppression was not present. This argument is a nonstarter. While Cassady was civil in nature, we “conclude[d] that the severability doctrine would not apply [t]here even if th[e] appeal were from a criminal suppression hearing.” Id. at 637. Second, the Government contends that the officers in this case did not treat the residential search
23
warrant as a general one because their search was limited to firearm-and-vehicle- related evidence. Even if that’s true, the point is irrelevant for the purpose of the “greater part” inquiry. That inquiry turns on what the warrant itself authorized, not “the extent of the actual search or the number of items seized.” Sells, 463 F.3d at 1159. And here, the warrant empowered officers to search for evidence of any crime. The warrant is therefore both invalid and non-severable.
C.
We have concluded that the residential search warrant violated the Fourth Amendment’s particularity requirement and cannot be severed. Yet the question of remedy remains.
1.
In general, evidence “obtained as a direct result of an unconstitutional search or seizure is plainly subject to exclusion.” Segura v. United States, 468 U.S. 796, 804 (1984). But as with most general rules, this one has a few exceptions. The Government urges us to apply one of them here: the good-faith exception to the exclusionary rule.
The district court initially said it would not address the good-faith exception because it found the residential search warrant sufficiently particularized. But later the court summarily concluded that even if the warrant was invalid, Officer Tomczyk reasonably relied on it when she conducted the initial sweep of Defendant’s home and saw the guns in the SUV parked under the carport. Because that’s all the district court said about the subject, the reasoning behind its conclusion is unclear.
24
It may be within our discretion to affirm the denial of Defendant’s suppression motion based on the good-faith exception. See United States v. Mosley, 743 F.3d 1317, 1324 n.2 (10th Cir. 2014). But it is no less an option for us to remand for the district court to fully address the issue in the first instance. And in this case, we think that is the more prudent course.
The inquiry on the good-faith exception requires not only an examination of the warrant’s text but also a careful consideration of the totality of the circumstances to determine whether officers reasonably relied on the invalid warrant. Russian, 848 F.3d at 1244, 1246. And here, the parties disagree on some potentially material aspects of the inquiry, such as whether officers limited their search to evidence only related to the vehicle shooting. The parties also dispute whether Officers Keith Wrede and Aaron Lloyd, who assisted Officer Menter search Defendant’s home, either read the warrant or reviewed any of its supporting documents before they executed the search. It’s also unclear from the record whether Officer Wrede or Officer Lloyd were otherwise informed of the warrant’s contents or briefed on what items to look for during the search.
The district court noted that Officer Menter’s testimony on those subjects was scattered (neither Officer Wrede nor Officer Lloyd testified), but it did not make any relevant factual findings. In these circumstances, we think it wise to allow the district court to do so given its superior resources for factfinding and the fact- intensive nature of the good-faith inquiry. See United States v. Gaines, 918 F.3d 793, 802–03 (10th Cir. 2019) (remanding for the district court to resolve factual disputes
25
and decide reasonable suspicion in the first instance); United States v. Nelson, 868 F.3d 885, 892 (10th Cir. 2017) (declining to consider whether the good-faith exception applied “because the record lack[ed] factual development on key issues”).
Even if the district court declines to take additional evidence on remand, the good-faith question is close on the current record. And before issuing a definitive decision, this court would benefit from a district court judgment that addresses the implications of previously unaddressed facts. In particular, is it fair to say Officer Tomczyk reasonably relied on the residential search warrant when she testified that she never received a copy of the warrant or reviewed Officer Menter’s affidavit? Maybe the answer is yes, since Officer Tomczyk was on Defendant’s property only to conduct a protective sweep, not to search for any items. Or perhaps Officer Tomczyk relied on the warrant in good faith because her supervisor, Sergeant Wolf, had reviewed the entire warrant package and directed her to lead the SWAT team in his stead. Then again, maybe not, as the warrant was the supposed lawful basis for Officer Tomczyk’s entry into Defendant’s property.
We won’t dwell on the point further, except to say that the good-faith question should get the full vetting it deserves so this court ultimately might be in a position to offer a judgment with the degree of confidence the question merits. Mindful that we are “a court of review, not of first view,” Cutter v. Wilkinson, 544 U.S. 709, 718 n.7 (2005), we will remand for the district court to determine whether the good-faith exception saves the incriminating evidence against Defendant from suppression.
26
2.
But wait, says the Government. It argues that even if the district court erred by admitting evidence discovered in Defendant’s home, the guns and ammunition found in the SUV are sufficient alone to uphold his conviction. As the Government correctly points out, Fourth Amendment violations, like most constitutional violations, “do not require reversal of a conviction if the court concludes the error was harmless beyond a reasonable doubt.” Russian, 848 F.3d at 1248. To apply that rule here, however, the evidence found in the SUV must have at least been admissible. See id. This is where the Government’s argument runs into a roadblock.
The exclusionary rule generally applied in Fourth Amendment cases requires courts to suppress not only evidence obtained as a “direct result of an illegal search or seizure” but also “evidence later discovered and found to be derivative of an illegality.” Utah v. Strieff, 136 S. Ct. 2056, 2061 (2016) (quoting Segura, 468 U.S. at 804). This so-called “fruit of the poisonous tree” doctrine does not demand a particularly tight causal chain between the illegal search and the discovery of the evidence sought to be suppressed. To satisfy the threshold causation requirement, a defendant need only show that “the evidence sought to be suppressed would not have come to light but for the government’s unconstitutional conduct.” United States v. Nava-Ramirez, 210 F.3d 1128, 1131 (10th Cir. 2000).
But-for causation, however, is only a necessary condition for suppression; it is not a guarantee the incriminating evidence will be suppressed. United States v. Shrum, 908 F.3d 1219, 1233 (10th Cir. 2018). Once a defendant establishes but-for
27
causation, the Government may avoid suppression by breaking the causal chain. To do so, the Government must show that “the incriminating evidence was discovered by means sufficiently distinguishable from the initial illegality to be purged of the primary taint.” Id.
Here, Defendant has established the requisite causal chain between illegal police conduct and the incriminating evidence he seeks to suppress. But for the invalid residential search warrant, Officer Tomczyk would not have entered the curtilage of Defendant’s home and seen the guns in the SUV parked under his carport. See Collins v. Virginia, 138 S. Ct. 1663, 1670–71 (2018) (holding that a partially enclosed carport constituted curtilage of the defendant’s home). And had Officer Tomczyk not seen those guns during the unlawful invasion of the home’s curtilage, see id., Officer Menter would not have used that information to procure the vehicle warrant, which led directly to the seizure of the evidence from the SUV.
The Government has made little effort to break the causal chain between the unconstitutional conduct and the evidence found in the SUV. It has not invoked the “attenuation” doctrine, which would allow it to avoid suppression by showing that “the connection between unconstitutional police conduct and the evidence is remote.” Strieff, 136 S. Ct. at 2061. Nor has the Government argued that the guns and ammunition found in the SUV would have been discovered even without the constitutional violation. See Nix v. Williams, 467 U.S. 431, 444 (1984) (adopting the inevitable-discovery doctrine).
28
In addition, the Government has not pointed to an independent source—that is, “a source wholly separate” from the initial unconstitutional intrusion—for the discovery and seizure of the evidence. Shrum, 908 F.3d at 1235. Although Officer Menter obtained a second warrant to search the SUV, he used information obtained during the illegal entry into the curtilage of Defendant’s home to secure the warrant. The Government has not explained how, under these circumstances, the warrant to search the SUV could serve as an independent source that purged the evidence of the primary taint arising from the unlawful entry. See Segura, 468 U.S. at 814 (applying the independent-source doctrine because “[n]one of the information on which the warrant was secured was derived from or related in any way to the initial entry into petitioners’ apartment”). And because it is the Government’s burden to show the independent-source doctrine applies, its failure to do so prevents us from upholding the denial of the suppression motion on that theory. Shrum, 908 F.3d at 1235.
The Government’s remaining arguments to avoid suppression of the evidence found in the SUV assume either the validity of the residential search warrant—a faulty premise—or that Officer Tomczyk relied on the invalid warrant in good faith, which has yet to be determined. So, at this point, we cannot conclude that any error in denying Defendant’s motion to suppress the evidence found in his home was harmless beyond a reasonable doubt.
IV.
For these reasons, we VACATE the denial of Defendant’s suppression motion and REMAND for the district court to determine whether the good-faith exception to
29
the exclusionary rule saves the incriminating evidence against Defendant from suppression.