dissenting:

I respectfully dissent, because shoplifting is not “similar to” any of the excluded offenses listed in U.S.S.G. § 4A1.2(c)(l). That being so, the district court did not err in counting Defendant’s prior conviction for shoplifting in his criminal-history score when sentencing him for the present crime of conviction. The majority’s contrary conclusion illustrates the inconsistencies in our cases, fails to follow the most closely analogous Ninth Circuit precedent, and needlessly creates a conflict with at least one sister circuit.

I. Misdemeanors generally are included in a criminal-history score.

At heart, U.S.S.G. § 4A1.2(c) is a rule of inclusion, not exclusion. It provides: “Sentences for misdemeanor and petty offenses are counted, except as follows.” What “follows” are two lists of excluded offenses, the (c)(1) list and the (c)(2) list.

The (c)(2) list is a list of crimes that “are of such minor significance to the goals of sentencing that inclusion would more likely distort than improve the process established by the guidelines for determining an appropriate sentence.” United States v. Martinez (Clyde), 905 F.2d 251, 253 (9th Cir.1990) (citation omitted). Those crimes of “minor significance” are hitchhiking, loitering, juvenile status offenses and truancy, public intoxication, vagrancy, and minor traffic infractions. U.S.S.G. § 4A1.2(e)(2). Under (c)(2), those offenses “and offenses similar to them, by whatever name they are known, are never counted ” toward a defendant’s criminal-history score. Id. (emphasis added).

The (c)(1) list, by contrast, is a list of crimes that sometimes are counted in a criminal-history score. Crimes on the (c)(1) list, “and offenses similar to them, by whatever name they are known,” are excluded from a defendant’s criminal-history score only if (a) the defendant received a light sentence and (b) the crime for which the defendant currently is being sentenced is not similar to the prior crime. U.S.S.G. § 4A1.2(c)(l). The crimes on the (c)(1) list include driving without a license, trespassing, non-support, prostitution, resisting arrest, contempt of court, and “[^Insufficient funds check.” The fact that the (c)(2) crimes never are counted, while the (c)(1) crimes sometimes are counted, reflects the Sentencing Commission’s judgment that the (c)(1) crimes are somewhat more serious than the crimes of “minor significance” on the (c)(2) list.

II. Under Nevada law, shoplifting is pet-it larceny.

Defendant pleaded guilty to “shoplifting” under Reno Municipal Code § 8.10.045 (1998). At the time, shoplifting was a subset of petit larceny (essentially, petit larceny in a store), the elements of which were set out in Reno Municipal Code § 8.10.040 (1998). As relevant, § 8.10.045 made it unlawful for “[a]ny person willfully to take possession of any merchandise offered for sale by any store with the intention of converting the same to the use of such person without paying to the owner the value thereof.” In 1999, the municipal code was amended; shoplifting merged into petit larceny and was eliminated as a separate crime.

As the majority notes, majority op. at 1026-27 n.l, it is of no consequence that Defendant was prosecuted under a local ordinance, because Defendant’s crime also was a crime under state law. Under Nevada Revised Statute § 205.0832, the crime that Defendant committed is punishable as theft. Theft, in turn, is a “single offense embracing certain separate of *1032 fenses,” including insufficient funds check and petit larceny. Nev.Rev.Stat. § 205.0833. A person commits petit larceny if he or she “[intentionally steals, takes and carries away ... or drives away” property of another that is worth less than $250. Nev.Rev.Stat. § 205.240.

So Defendant’s local conviction for shoplifting was equivalent to a state conviction for theft or, more specifically, for petit larceny. Like other forms of petit larceny, shoplifting required an attempted or completed trespassory taking of property, with the intent permanently to deprive the owner of that property. The acts covered by the municipal code are completely encompassed in the state law against larceny.

III. This circuit has adopted two tests for determining ivhen one crime is “similar to” another.

Petit larceny (or theft, or shoplifting) is not among the crimes on either the (c)(1) or the (c)(2) list. But Defendant argues— and the majority concludes — that Defendant’s conviction cannot be counted because it is “similar to” a crime on the (c)(1) list, “insufficient funds check.”

The majority acknowledges, majority op. at 1027-28, that this court has two tests for determining whether one crime is “similar to” another under U.S.S.G. § 4A1.2(c): the Martinez (Clyde) test and the Martinez (Carlos) 1 test. The majority then applies both tests and concludes that, under either, petit larceny is “similar to” insufficient funds check. I disagree with the majority’s application of the two Martinez tests; but before explaining my disagreement, I wish to explain why the Martinez (Clyde) test is neither useful nor appropriate in cases like this one.

This case involves application of U.S.S.G. § 4A1.2(c)(l). But in Martinez (Clyde) , this court was applying U.S.S.G. § 4A1.2(c)(2). As noted, (c)(2) lists the least significant offenses of which a defendant can be convicted — offenses such as minor traffic infractions. The question in MaHinez (Clyde) was how to decide whether the offense at issue, public indecency (which was not on that list), was nevertheless of “such minor significance” that it never should be counted for purposes of sentencing.

In making that determination, the Martinez (Clyde) court took a “categorical,” rather than a “crime-by-crime,” approach. By that I mean that the court did not pick a particular crime from the (c)(2) list and then determine whether public indecency was similar to it. Rather, the court described the general characteristics of the “insignificant” offenses on the (c)(2) list as a whole and then set about to determine whether public indecency was an offense that shared those general characteristics. The court identified three such characteristics, asking whether: (1) the offense “of'fer[s] [a] basis for predicting future significant criminal activity by the defendant”; (2) the offense involves conduct that “is not uniformly criminalized, and when it is, the penalty is usually light”; and (3) “[sjociety has a substantial interest in punishing such conduct ... and in discouraging its repetition.” MaHinez (Clyde), 905 F.2d at 253-54 ; see also United States v. Sandoval, 152 F.3d 1190, 1192 (9th Cir.1998) (listing elements). 2 In particular, the court fo *1033 cused on the fact that the offenses on the (c)(2) list are not universally regarded as culpable: “Consideration of convictions for such conduct in determining sentences would be inimical to the realization of the goal of uniformity in sentencing since the sentencing calculation under the guidelines would vary with the jurisdiction in which a person who engaged in such conduct resided.” Ma rtinez (Clyde), 905 F.2d at 253 .

The court then addressed the crime of public indecency in the light of those general characteristics and concluded that it was not “similar to” the crimes on the (c)(2) list. Id. at 254 . The court focused on the fact that public indecency is generally regarded as culpable, as evidenced by the fact that it is treated as a criminal offense in every American jurisdiction. Id. “Because public indecency is universally regarded as culpable,” the court concluded, “conviction of such an offense is relevant to the likelihood the offender will engage in criminal conduct in the future and is significant for sentencing purposes.” Id.

Like Martinez (Clyde), United States v. Martinez (Carlos), 69 F.3d 999 (9th Cir.1995), is a case that arose under (c)(2). But instead of following the categorical “general characteristics” approach espoused in MaHinez (Clyde), the Martinez (Carlos) court looked at the specific conduct underlying the defendant’s prior offense, vandalism. Id. at 1000 . The court reasoned: “The inquiry into whether vandalism is ‘similar’ to a section 4A1.2(c)(2) listed offense must focus on whether the activity underlying vandalism is similar to the activities underlying the listed offenses.” Id. The court then considered whether the “activity underlying” certain offenses on the (c)(2) list was similar to the conduct underlying vandalism and concluded that it was not. Id. at 1000-01 . Accordingly, the court held that none of the offenses on the (c)(2) list was “similar to” vandalism. In explaining its refusal to follow the categorical MaHinez (Clyde) test, the court stated that application of that test “would lead to the same result, but it would do so without looking at the real question before this couH — the similarity of the prior offense of vandalism to the listed excluded offenses.” Id. at 1001 (emphasis added).

The MaHinez (Clyde) test and the Martinez (Curios) test answer related but different questions. The MaHinez (Clyde) test answers this question: “As a general matter, is this crime the kind of crime that the Sentencing Commission placed on the (c)(2) list?” The MaHinez (Carlos) test answers this question: “Is this crime similar to any paHicular crime on the (c)(2) list?”

IV. In the context of § 4.Al.2(c)(l), only the MaHinez (Carlos) test should apply.

Our question in this ease is the MaHinez (Carlos) question, albeit in the context of (c)(1). We must answer, not the question whether petit larceny is like the crimes on the (c)(1) list as a general matter but, rather, the question whether petit larceny is like a specific crime on the (c)(1) list, insufficient funds check.

The MaHinez (Clyde) test is ill-suited to answering that specific question, paHicu-larly when it is applied, as here, to a case under (c)(1). 3 A review of the three elements of the test illustrates why. First, Martinez (Clyde) instructs us to ask whether a crime is uniformly considered to be culpable. 905 F.2d at 253 . The offenses on the (c)(2) list are not criminalized *1034 in all jurisdictions. Thus, an offense that is considered a crime in some places but not in others is, in that respect, similar to the crimes on the (c)(2) list. It would be unfair, and contrary to the goal of national uniformity, ever to count such an offense in a criminal-history score.

But there is no suggestion that the more significant crimes on the (c)(1) list are not uniformly considered to be culpable. Certainly the offenses at issue here are criminalized, in one form or another, in every jurisdiction. See Model Penal Code (1980): § 223.2 (theft by unlawful taking); § 224.5 (bad checks). In that sense, I suppose that the crimes are similar; both are uniformly considered to be culpable. But it is nonsensical to conclude that such a similarity is a reason not to count a crime for criminal-history purposes. So the first question from Martinez (Clyde) is simply unhelpful in cases that address the (c)(1) crimes, which are uniformly regarded to be culpable to some degree.

Second, Martinez (Clyde) instructs us to consider whether a crime offers a “basis for predicting future significant criminal activity by the defendant.” 905 F.2d at 253 . Again, that question makes sense in a categorical, Martinez (Clyde) inquiry, in which a court is trying to determine whether a crime is in the class of crimes that are so minor that they never should be counted. But here, we are trying to determine whether a specific crime, petit larceny, is like another specific crime, insufficient funds check. How could our speculation about whether thieves are likely to commit crimes again help us answer that question? If we conclude that they are, does that make them more “similar to” people who pass bad checks, or less so? Is insufficient funds check a crime that is characterized by a likelihood of recidivism? How are we to know that? The majority implicitly acknowledges that this step in the Martinez (Clyde) test is impossible to apply here by listing the step, majority op. at 1028, but never discussing it.

Third, Martinez (Clyde) instructs us to consider whether “[sjociety has a substantial interest in punishing such conduct ... and in discouraging its repetition.” 905 F.2d at 254 . As with the other parts of the test, this question is difficult to apply outside the context of a categorical, Martinez (Clyde) inquiry. Every jurisdiction punishes petit larceny and insufficient funds check (just as, with minor exceptions, they punish the rest of the crimes on the (c)(1) list). So in that sense the two crimes are “similar.” But again, why should the fact that both crimes are condemned everywhere argue against counting petit larceny for purposes of criminal history?

The majority also considers another question in its analysis under Martinez (Clyde) — Defendant’s sentence. As the majority notes, there is some disagreement in this circuit about whether that is an appropriate consideration under (c)(2). Majority op. at 1029 n.5. But there should be no disagreement that such a consideration is inappropriate under (c)(1). That is because (c)(1), unlike (c)(2), already contains within it a sentencing consideration: the threshold requirement that a prior offense may be excluded only if the term of punishment actually imposed was less than 30 days in prison or one year’s probation. In other words, (c)(1) already considers sentencing; even to reach the question whether a crime is “similar to” a crime on the (c)(1) list, a court must first satisfy itself that the sentencing criteria are met. Accordingly, it is unnecessary under (c)(1) to examine whether a defendant received a light sentence in determining whether two crimes are “similar.” The answer must always be “yes.” United States v. Harris, 128 F.3d 850, 855 (4th Cir.1997).

In sum, the Martinez (Clyde) test is inappropriate and unhelpful in this case. In cases under (c)(1), where the inquiry is whether a specific crime is similar to another specific crime, the better approach is that taken in Martinez (Carlos) . A court simply should look at the two crimes that are alleged to be similar and “focus on whether the activity underlying [one] is *1035 similar to the activity] underlying the [other].” Martinez (Carlos), 69 F.3d at 1000 .

V. Under either test, these crimes are not “similar.”

A. The Martinez (Carlos) test

The majority concludes that Defendant’s theft is similar to insufficient funds check under the Martinez (Carlos) test, largely because the crimes have several overlapping elements. I disagree. As noted, the offense of “shoplifting” under the 1998 Reno Municipal Code, and the offense of “petit larceny” under Nevada state law, require a trespassory taking of property. It is that element that differentiates larceny from fraud-based offenses like passing a bad check. Neither insufficient funds check nor any other offense that is excluded under U.S.S.G. § 4Al(2)(c) pertains to a trespassory taking of the property of another.

As support for its contrary conclusion, the majority cites United States v. Sandoval. Majority op. at 1030. In Sandoval , this court held that petty theft was not similar to any of the offenses on the (c)(2) list, primarily because none of the offenses on the (c)(2) list requires proof of criminal intent or involves stealing. 152 F.3d at 1192 . From this, the majority concludes that petit larceny and insufficient funds check must be similar, because both require criminal intent and stealing.

Sandoval is this court’s only case addressing the question whether a minor theft crime like petit larceny is excludable under U.S.S.G. § 4A1.2(c). The Sandoval court concluded that the crime was not excludable. 152 F.3d at 1192 . The court’s holding, therefore, would appear to support the argument that petit larceny should not be excluded in this case. But the majority cites it for the opposite proposition. To be sure, Sandoval may be distinguished on the ground that it is a(c)(2) case, not a(e)(l) case like this one. But so are the other cases on which the majority bases its holding, Martinez (Clyde) and Martinez (Carlos) .

In any event, Sandoval does not support the majority’s holding. Sandoval simply applies Martinez (Carlos) and concludes that, if there is some significant difference between the conduct underlying a prior misdemeanor and the conduct underlying a listed offense, then the two crimes are not similar. It is true that petit larceny is not different from insufficient funds check in all respects. Petit larceny and insufficient funds check both differ from the crimes listed in (e)(2) for the reasons discussed above in Part IV. But petit larceny and insufficient funds check nevertheless differ from each other, and significantly so; and Martinez (Carlos) requires no more. Pet-it larceny is like insufficient funds check in that both involve stealing, but it is different in that it involves a trespassory taking, or “physical possession or control without consent or authority.” Model Penal Code § 223.2, at 165 (1980). That sort of physical taking without consent is simply different from the act of obtaining property by fraud. This additional element distinguishes the activity underlying the two crimes for purposes of the Martinez (Carlos) test.

B. The Martinez (Clyde) test

Nor are these offenses similar under the Martinez (Clyde) test, even assuming that it applies. As noted, petit larceny is criminalized universally. Martinez (Clyde), 905 F.2d at 253 . “Society has a substantial interest in punishing such conduct ... and in discouraging its repetition.” Id. at 254 . And, under the reasoning of Martinez (Clyde) , petit larceny is predictive of future criminal activity. In Martinez (Clyde) , the court addressed this step through reference to the universal culpability of the offense in question: “Because public indecency is universally regarded as culpable, conviction of such an offense is relevant to the likelihood the offender will engage in criminal conduct in the future and is significant for sentencing purposes.” Id. So too here.

*1036 The majority also addresses the elements of the offenses in its discussion of the Martinez (Clyde) test. Majority op. at 1028-29. Assuming that this properly is part of the Martinez (Clyde) inquiry, but see ante note 2, I do not agree that the elements of these offenses support the conclusion that petit larceny should be excluded under (c)(1). Although the offenses have some elements in common, they are crucially different, as discussed above.

VI. The majority creates an inter-circuit conflict.

Every circuit to have considered this question has held that petit larceny (or petty theft, or shoplifting) is not “similar to” to any offense on the (c)(1) list, including insufficient funds check. United States v. Waller, 218 F.3d 856, 857-58 (8th Cir.2000); United States v. Hooks, 65 F.3d 850, 855 (10th Cir.1995); United States v. Ziglin, 964 F.2d 756, 758 (8th Cir.1992); United States v. Hoelscher, 914 F.2d 1527, 1537 (8th Cir.1990); see also United States v. Benjamin, 110 F.3d 61 , 1997 WL 165414 (4th Cir.1997) (unpublished disposition); United States v. Mazanetz, 61 F.3d 906 , 1995 WL 431404 (7th Cir.1995) (unpublished disposition). 4 With this opinion, we become the first circuit court to hold otherwise.

The majority asserts that its creation of a circuit split may not last long, because the Eighth Circuit might see fit to reconsider its binding precedent now that “insufficient funds check” is on the (c)(1) list. Majority op. at 1030. A change of course seems unlikely; the Eighth Circuit specifically has rejected a defendant’s argument “that he should not have received one criminal history point for a prior state petty-theft conviction, because it is similar to the offense of ‘[ijnsufficient funds check,’ which is included on an exclusionary list in U.S. Sentencing Guidelines Manual § 4A1.2(c)(l) (1997).” United States v. Tauer, 141 F.3d 1171 , 1998 WL 85348, at *1 (8th Cir.1998) (unpublished disposition). 5 Mazanetz, likewise, rejected the very argument that Defendant makes here. See also United States v. Dershem, 818 F.Supp. 785, 791 (M.D.Pa.1993) (holding that “retail theft” is not excludable under U.S.S.G. § 4Al.2(c)(l), in response to the defendant’s argument that it is analogous to “insufficient funds check”). The majority also downplays the creation of a circuit split by pointing out distinguishing features of some the cases cited above. Majority op. at 1030-31.

But the fact remains that, after this opinion, the Ninth Circuit will stand alone, and sentencing disparities will occur for the first time on this basis. In this circuit, and this circuit only, defendants’ prior convictions for petty theft will be excluded under U.S.S.G. § 4A1.2(c). It is striking that the case on which the majority most heavily relies, Martinez (Clyde) , based its holding on the “realization of the goal of uniformity in sentencing” and on the necessity of avoiding a situation in which “sentence calculation[s] under the guidelines would vary with the jurisdiction in which a person who engaged in such conduct resided.” 905 F.2d at 253 .

VII. The majority’s result is contrary to the likely intent of the Sentencing Commission.

I disagree with the majority for one final reason: its result simply does not make *1037 sense. Larceny (or theft) is by far the most frequently committed crime that the FBI lists in its crime index. See 1999 FBI Crime Index at Figure 2.3. In 1999, larceny and theft made up almost 60% of the total crimes reported in the crime index. Id. In that year alone, the FBI reported nearly 7 million larcenies and thefts. Id. § II.

Petit larceny (or shoplifting, or petty theft) assuredly is one of the most common, and best known, of misdemeanors. The majority concludes that the Sentencing Commission intended this crime to be excluded from criminal-history scores under § 4A1.2(c)(l). If so, then why did not the Commission simply put this crime on the (c)(1) list? There are 15 crimes on the list, including such less obvious (and less common) misdemeanors as hindering a police officer and fish and game violations. Is it really likely that the Commission intended to list petit larceny — perhaps the prototypical misdemeanor — but simply neglected to do so, or thought that it was unnecessary because petit larceny is so clearly “similar to” insufficient funds check?

Of course not. If the Commission had wanted to list this offense, it would have done so. But it did not.

VIII. Conclusion

For the reasons stated, I disagree with the majority’s conclusion that petit larceny is “similar to” insufficient funds check for purposes of U.S.S.G. § 4A1.2(c)(l). I therefore dissent.

. United States v. Martinez (Carlos), 69 F.3d 999 (9th Cir.1995).

. The majority insists that my reading of Martinez (Clyde) is unduly restrictive. Majority op. at 1028. My description of the three-part test, however, mirrors the numbered description of the "three-factor test" of Martinez (Clyde) that this court recently gave in Sandoval, 152 F.3d at 1192 .

The majority cites United States v. Kemp, 938 F.2d 1020, 1023 (9th Cir.1991), for the proposition that the “Martinez (Clyde) analysis includes a comparison of the elements of the listed offense and the prior offense,” which would be a fourth factor and would make this court's test similar to the Fifth Circuit's formulation in United States v. Hardeman, 933 F.2d 278, 281 (5th Cir.1991). Majority op. at 1028. Assuming that Kemp contains such a suggestion, the difference between Kemp and Sandoval only highlights the disarray in our precedents. Now we apparently have three tests, rather than two: Martinez (Clyde) as interpreted and applied in Sandoval; Martinez (Clyde) as interpreted and applied in this case and, possibly, in Kemp ; and Martinez *1033 (Carlos). Al least one other circuit has criticized the Ninth Circuit's law with respect to U.S.S.G. § 4A1.2(c) as "indeterminate” and as creating "confusion.” United States v. Harris, 128 F.3d 850, 854 (4th Cir.1997).

. This court has attempted, only once, to apply the Martinez (Clyde) test in the context of (c)(1). Kemp, 938 F.2d 1020 . The court concluded that it was "practically impossible” to apply the test in the circumstances of the case and, accordingly, "carve[d] out an exception” to it by looking at the underlying conduct involved in the prior crime. Id. at 1023 .

. The Fourth Circuit allows citation of unpublished opinions if “there is no published opinion that would serve as well.” United States Court of Appeals for the Fourth Circuit, Rule 36(c). The Seventh Circuit does not. United States Court of Appeals for the Seventh Circuit, Rule 53(2)(iv). I do not cite these dispositions for their precedential or persuasive value but, rather, simply to note their existence.

. In the Eighth Circuit, an unpublished opinion may be cited if it “has persuasive value on a material issue and no published opinion of this or another court would serve as well.” United States Court of Appeals for the Eighth Circuit, Rule 28A(i); see also Anastasoff v. United States, 235 F.3d 1054, 1056 (8th Cir.2000) (stating that the constitutionality of the provision in Rule 28A(i) that unpublished opinions have no precedential effect "remains an open question” in that circuit).