(dissenting).
I respectfully dissent. I disagree with the majority’s conclusion that the sentencing disparity between Minn.Stat. § 609.535, subd. 2a(a)(l) (2010), and Minn. Stat. § 609.52, subd. 3(4) (2010), is constitutional as applied to the appellant, Diane Marie Cox. Laws that purport to address social goals that the Legislature believes to be important must have a rational basis, or else the Legislature runs the risk of having the law overturned because equal protection rights are violated. Here, I conclude that Minnesota’s statutory scheme for punishing the practice of issuing dishonored checks does not have a rational basis as applied to Cox’s actions.
Further, in reaching its decision to uphold Minnesota’s statutory scheme as applied to Cox in this case, I conclude that the majority fails to resolve the inconsistencies in our equal protection case law, and incorrectly applies the test that it chooses to utilize. To resolve these inconsistencies, I would hold that the “similarly-situated” part of our rational basis review utilizes a threshold test, and that our three-part rational basis test applies once this threshold has been crossed. After applying these tests to the facts of this case, I would hold that Cox has not only met the threshold test, but has also established that the sentencing disparity between section 609.535 and section 609.52 does not satisfy our three-part rational basis test. Therefore, I would reverse the court of appeals.
Before articulating the legal analysis that underlies the reasons for my dissent, a brief review of some of the relevant facts of this case is in order. Diane Marie Cox is accused of issuing dishonored checks worth a total of $515.83 in violation of Minn.Stat. § 609.535, subd. 2 (2010). If convicted of this offense, she faces a maximum penalty of five years in prison, a $10,000 fine, and being labeled a felon for the rest of her life. Minn.Stat. § 609.535, subd. 2a(a)(l) (2010). Had Cox been charged with the more serious offense of theft by check11 — an offense requiring the *527State to prove that she had a higher degree of mens rea and that she obtained property or services from another in exchange for the dishonored checks — she would, if convicted, face a maximum penalty of one year in jail, a $3,000 fíne, and have only a gross misdemeanor on her record. Minn.Stat. § 609.52, subds. 2(3)(i), 3(4) (2010). In essence, because Cox was charged with the less-serious crime — that is, the less-culpable and less-blameworthy crime — she faces a potentially greater punishment.
The State fails to justify this illogical punishment disparity, stating that the disparity may have been simply “inadvertence” by the Legislature in amending one of the statutes but not the other. But this alleged “inadvertence” has real-life consequences for Cox. It could result in a four year greater prison sentence, $7,000 more in fines, and impose upon her the label of a felon. Such a markedly different result does not have a rational basis.
I.
I readily acknowledge that the development of equal protection case law by our court has been marked by some inconsistency. It also appears that these inconsistencies are being perpetuated by the majority’s willingness to apply a test it derives from some of our cases, while at the same time, the majority either summarily dismisses or ignores several of our cases that dictate a different result.
The majority asserts that it is applying the similarly-situated test to deny Cox’s equal protection challenge. We have previously used this similarly-situated test in several cases to deny equal protection claims. In State v. Witt, we stated that “[t]he sole limitation which the equal protection clause imposes upon the legislature in the exercise of this power is that criminal statutes must not prescribe different punishments for the same acts committed under the same circumstances by persons in like situation.” 310 Minn. 211, 215, 245 N.W.2d 612, 616 (1976) (citation omitted) (internal quotation marks omitted). We have repeatedly relied on comparable formulations of the similarly-situated test. In State v. Frazier we stated that “[a] statute violates the equal protection clause when it prescribes different punishments or different degrees of punishment for the same conduct committed under the same circumstances by persons similarly situated.” 649 N.W.2d 828, 837 (Minn.2002). Earlier, in State v. Dietz, we stated that “[ejqual protection of the laws is not denied by a statute prescribing the punishment to be inflicted on a person convicted of crime, unless it prescribes different punishments for the same acts committed under the same circumstances by persons in like situation.” 264 Minn. 551, 558, 119 N.W.2d 833, 837 (1963) (citation omitted) (internal quotation marks omitted).
We have, however, decided other equal protection challenges by applying a more robust test — a test that we have applied in equal protection cases since at least 1979. In these other cases, we have said that to satisfy rational basis review under an equal protection challenge pursuant to the Minnesota Constitution, a statute must meet each of the following parts of a three-part test:
(1) The distinctions which separate those included within the classification from those excluded must not be manifestly arbitrary or fanciful but must be genuine and substantial, thereby providing a natural and reasonable basis to justify legislation adapted to peculiar conditions and needs; (2) the classification must be genuine or relevant to the *528purpose of the law; that is, there must be an evident connection between the distinctive needs peculiar to the class and the prescribed remedy; (3) the purpose of the statute must be one that the state can legitimately attempt to achieve.
Miller Brewing Co. v. State, 284 N.W.2d 353, 356 (Minn.1979); see also State v. Garcia, 683 N.W.2d 294, 299 (Minn.2004) (applying the three-part test); State v. Benniefield, 678 N.W.2d 42, 46 (Minn.2004) (same); State v. Russell, 477 N.W.2d 886, 888 (Minn.1991) (same).12
When applying the foregoing three-part test, we have said that we have “been unwilling to hypothesize a rational basis to justify a classification.... Instead, we have required a reasonable connection between the actual, and not just the theoretical, effect of the challenged classification and the statutory goals.” Russell, 477 N.W.2d at 889. We have said that we have been “consistent” in applying this aspect of the three-part test, id., and that this aspect was “[t]he key distinction between” the three-part test and the federal rational basis test. Garcia, 683 N.W.2d at 299. But, a thorough review of our case law shows just the opposite — there is a lack of consistency in applying the no-hypothetical-justifications aspect of our three-part test. For example, in Bennie-field, we made no mention of the requirement that we analyze the actual justification for the classification. See 678 N.W.2d at 46-47; see also State v. Barnes, 713 N.W.2d 325, 332-33 (Minn.2006) (making no mention of a requirement that the court only consider the actual justification for the classification).
To our credit, we have acknowledged that we have been inconsistent in applying the three-part test outlined above. In Russell, we stated that we have “not been consistent in explaining whether the rational basis standard under Minnesota law, although articulated differently, is identical to the federal standard or represents a less deferential standard under the Minnesota Constitution.” 477 N.W.2d at 889. We noted that the federal standard for rational basis review requires “(1) a legitimate purpose for the challenged legislation, and (2) that it was reasonable for the lawmakers to believe that use of the challenged classification would promote that purpose.” Id. at 887. Unfortunately, acknowledging inconsistencies, but not resolving them, is not helpful to courts relying on us for guidance.
For more than twenty-five years, we have recognized inconsistencies in our *529equal protection case law, but have failed to resolve them. In 1985, we recognized that our three-part test for rational basis review may be a stricter standard than the federal rational basis standard and said the issue “deserves our careful consideration.” Bernthal v. City of St. Paul, 376 N.W.2d 422, 424-25 (Minn.1985). But in Bemthal, we concluded that we should defer further analysis because resolution of the issue was not dispositive. Id. at 425. In 1993, Justice Esther Tomljanovich provided a “careful consideration,” id. at 425, of this issue, but in a dissenting opinion. See Mitchell v. Steffen, 504 N.W.2d 198, 210 (Minn.1993) (Tomljanovich, J., dissenting). Since then, our court has merely acknowledged difficulties in our case law, but has done so without actually resolving these difficulties. See Scott v. Minneapolis Police Relief Ass’n, Inc., 615 N.W.2d 66, 74 n. 15 (Minn.2000) (stating the three-part and federal rational basis tests “have raised questions concerning their applicability in different contexts,” but not answering those questions); see also Kolton v. Cnty. of Anoka, 645 N.W.2d 403, 412 n. 5 (Minn.2002) (acknowledging that the applicability of the three-part and federal rational basis tests is unclear). Given the history of our equal protection jurisprudence, I find it problematic that the majority’s analysis does not address why its similarly-situated test is the proper test as opposed to any of these other tests that we have also applied. I believe that we have an obligation to reconcile our case law so that Minnesota courts can apply a consistent equal protection standard. Therefore, I will attempt such a reconciliation.13
A.
As a starting point for this reconciliation, I conclude that the similarly-situated test applied by the majority to uphold the charge against Cox should be viewed as a threshold test. In other words, we should apply the similarly-situated test first to determine whether the challenge to the classification actually implicates equal protection rights. Only if the individual asserting a claim meets the similarly-situated test should the court ask whether the challenged classification can survive rational basis review. As the majority correctly notes, the similarly-situated test has not always been applied in equal protection cases. In both Barnes and Russell, we applied the three-part rational basis test without first applying the similarly-situated test. See Barnes, 713 N.W.2d at 332; Russell, 477 N.W.2d at 887-89. However, for the following reasons, I believe that treating the similarly-situated test as a threshold test is the correct way to approach an equal protection challenge.14
First, the general goal of the equal protection right is to ensure that similarly-situated individuals are treated similarly. In Scott, we stated that our equal protection analysis “begin[s] with the mandate that all similarly situated individuals shall be treated alike, but only ‘invidious discrimination’ is deemed constitutionally of*530fensive.” 615 N.W.2d at 74 (citation omitted). Similarly, in State by Spannaus v. Lutsen Resorts, Inc., we stated “[ejssential to a ruling that equal protection has been denied by discriminatory administration of the laws is a finding that the persons treated disparately are similarly situated.” 310 N.W.2d 495, 497 (Minn.1981). This goal of treating similarly-situated individuals similarly is mentioned in several cases applying the three-part test. See, e.g., Garcia, 683 N.W.2d at 298-99 (quoting the Scott language quoted above, but then outlining the three-prong test).
Second, when we apply the similarly-situated test to deny an equal protection challenge, we generally do not state the level of review we are applying. One step in deciding an equal protection challenge is to determine the applicable level of review. See Greene v. Comm’r of Minn. Dep’t of Human Servs., 755 N.W.2d 713, 725 (Minn.2008) (“We must determine whether to apply strict scrutiny or rational basis review to Greene’s equal protection claim.”). What level of review to apply will determine “the degree of deference to be accorded the legislative judgment when it creates classes treating persons within and without the class differently.” Mitchell, 504 N.W.2d at 210 (Tomljanovich, J., dissenting). Once we determine the level of review — rational basis, strict scrutiny, or an intermediate level of scrutiny — we apply that level of review to the facts of the case. See Greene, 755 N.W.2d at 729.
When we apply the similarly-situated test to deny an equal protection challenge, however, we do not determine which level of review to apply, nor do we apply a level of review.15 See, e.g., Hale v. Viking Trucking Co., 654 N.W.2d 119, 125-26 (Minn.2002) (denying a would-be interve-nor’s equal protection violation argument without applying any level of review); Frazier, 649 N.W.2d at 837-39 (denying appellant’s equal protection challenge without applying a level of review). This omission from our opinions indicates that the claim is being denied at a threshold level. Recently, Justice Alan Page agreed with this analysis in a concurring opinion, stating that “if the people are not similarly situated there is no equal protection claim in the first instance and there is no need to engage in a discussion about the appropriate standard of review and how that standard is to be applied.” Clayton v. Kiffmeyer, 688 N.W.2d 117, 133 (Minn.2004) (Page, J., concurring).
Third, my research has uncovered no case from our court in which we applied only the similarly-situated test and then found an equal protection violation. Threshold tests are useful only in denying claims that lack merit; threshold tests are not useful in determining the validity of a claim once the claim passes this threshold. In other words, once claimants meet the similarly-situated test and get past this threshold, courts need to apply some level of review to determine the merits of the claim. The lack of cases applying the similarly-situated test to find an equal protection violation supports the notion that the test is a threshold test.
Finally, the similarly-situated test has been used as a threshold test in at least one of our prior eases. In Garcia, we considered whether equal protection rights were denied by granting different jail time credits to juveniles certified as adults who violated probation and juveniles certified as extended jurisdiction juveniles who violated probation. 683 N.W.2d at 298. As part of our analysis, we outlined the vari*531ous rational basis tests we had used in the past, and decided we would use the three-part test. Id. at 298-99. Next, we stated that “[w]e first must consider the state’s argument that juveniles and adults are not similarly situated for equal protection purposes.” Id. at 299. Because we concluded that individuals in the two classifications were similarly situated, we proceeded to “consider whether, on the facts presented, there is a rational basis” for the distinction in granting jail credits. Id. We then applied the three-part test to determine whether the State’s classification met rational basis review. Id. at 299-301. In other words, in Garcia, we applied the similarly-situated test as a threshold test. Finding the threshold test satisfied, we then applied our three-part test used for rational basis review.
B.
As previously noted, there is our three-part rational basis test as well as a federal version of the rational basis test for equal protection challenges. Our current three-part test states the following:
(1) The distinctions which separate those included within the classification from those excluded must not be manifestly arbitrary or fanciful but must be genuine and substantial, thereby providing a natural and reasonable basis to justify legislation adapted to peculiar conditions and needs; (2) the classification must be genuine or relevant to the purpose of the law; that is there must be an evident connection between the distinctive needs peculiar to the class and the prescribed remedy; and (3) the purpose of the statute must be one that the state can legitimately attempt to achieve.
Russell, 477 N.W.2d at 888. The federal rational basis test states that “legislation is presumed to be valid and will be sustained if the classification drawn by the statute is rationally related to a legitimate state interest.” City of Cleburne v. Cleburne Living Ctr., 473 U.S. 432, 440, 105 S.Ct. 3249, 87 L.Ed.2d 313 (1985).
We have applied our three-part test to equal protection claims under the Minnesota Constitution since 1979.16 At times, however, we have applied the federal rational basis test to such claims. For example, in State v. Behl, we analyzed an equal protection challenge to a sentencing disparity based solely on whether a juvenile was indicted, but not convicted, of first-degree murder. 564 N.W.2d 560, 568-69 (Minn.1997). In Behl, we did not mention our three-part test when applying the federal rational basis test.17 Id. at 568; see also Lidberg v. Steffen, 514 N.W.2d 779, 784 (Minn.1994) (applying the federal rational basis test to an equal protection challenge for different commitment discharge procedures).
*532Justice Tomljanovich, in her dissenting opinion in Mitchell, suggested that the three-part test outlined in Russell be reserved as an intermediate level of review to be applied only when there is a semi-suspect class, or when a facially neutral statute has a disparate impact on a suspect class, as was the case in Russell. 504 N.W.2d at 210 (Tomljanovich, J., dissenting). To support this suggestion, Justice Tomljanovich pointed to three cases decided shortly after Russell that applied the federal rational basis test, which indicated to her that the three-part test was not our default test for rational basis review. Id. Since Justice Tomljanovieh’s dissent in Mitchell, however, we have applied the three-part test in several cases that do not implicate a semi-suspect class or a disparate impact on a suspect class. In Barnes, we applied the three-part test to an equal protection challenge based on disparate treatment under a domestic abuse murder statute. 713 N.W.2d at 332-33. In Benniefield, we applied the three-part test to a case challenging a sentencing disparity between possessing drugs inside a school zone and possessing drugs outside a school zone. 678 N.W.2d at 45-47. Also, as previously noted, we applied the three-part test in Garcia, a case involving different jail time credits given to juveniles certified as adults and juveniles certified as extended jurisdiction juveniles. 683 N.W.2d at 296, 298-301. These eases indicate that since Russell we have developed and applied our three-part test as the default test for rational basis review in equal protection cases under the Minnesota Constitution. Therefore, I conclude that based on our most relevant case law, we should apply the three-part test to equal protection challenges brought under the Minnesota Constitution that warrant rational basis review and meet the similarly-situated threshold test.18
C.
The next question to be addressed as part of my attempt to reconcile our equal protection case law is whether our three-part rational basis test allows us to consider hypothetical justifications for the classifications challenged in an equal protection case, or whether we must look exclusively to the Legislature’s actual justification(s). The federal test states that a statutory classification will survive rational basis review “if there is any reasonably conceivable state of facts that could provide a rational basis for the classification.” F.C.C. v. Beach Commc’ns, Inc., 508 U.S. 307, 313, 113 S.Ct. 2096, 124 L.Ed.2d 211 (1993). I would conclude that when applying our three-part rational basis test, we may look to hypothetical justifications for the statute, just as in the federal test.
Purported case law basis for the no-hypothetical-justifications requirement is unfounded
In Russell, we stated that we have “been consistent” in requiring that courts analyze “the actual, and not just the theoretical” justification for a challenged classification. 477 N.W.2d at 889. For support, we then cited to a William Mitchell Law Review article from 1984 that discusses three of our cases. Id. (citing Deborah K. McKnight, Minnesota Rational Relation Test: The Lochner Monster in the 10,000 Lakes, 10 Wm. Mitchell L.Rev. 709 (1984))19; see also In re Estate of Turner, *533391 N.W.2d 767, 772 (Minn.1986) (Wahl, J., concurring) (citing the same cases and article for the same proposition). I disagree that our cases require us to consider only the actual justification for the challenged statute.
In Russell, we first cited to Wegan v. Village of Lexington, 309 N.W.2d 273 (Minn.1981), for the proposition that we assess the actual justification for a challenged classification. Russell, 477 N.W.2d at 889. Wegan involved a disparity in the Dram Shop Act between the limitations period applicable to individuals suing retailers that sold beer with less than 3.2% alcohol (“non-intoxicating” liquor retailers), and all other liquor retailers (“intoxicating” liquor retailers). 309 N.W.2d at 277-78. A plaintiff suing an intoxicating liquor retailer for injuries had one year to commence an action under the Dram Shop Act, whereas a plaintiff suing a non-intoxicating liquor retailer under common law had six years to commence an action. Id. We held that this time disparity violated equal protection. Id. at 281.
In deciding Wegan, we applied our three-part test for rational basis review. Id. at 280. We concluded that there was a legitimate purpose to the statute (part 3 of our test), but concluded that the distinction between the two types of liquor retailers was arbitrary (part 1), and that the state’s classification of the retailers was not relevant to the purpose of the Dram Shop Act (part 2). Id. We then held the classification to be arbitrary because it did not matter which type of alcohol was used to intoxicate a person when a plaintiff was injured by an intoxicated person. Id. In Wegan, we did not discuss hypothetical justifications for the classification. See id. at 280-81. McKnight argues this lack of any discussion of hypothetical justifications for the statute shows a departure from the federal version of rational basis review because under our three-part test, we only considered the actual justification for the challenged statute. McKnight, supra, at 726. I believe this argument is, at best, a reach, and at worst, unfounded.
What I find important in our Wegan opinion is that we did not state that we were only considering the actual justification for the statute. Nor did we state that future courts may not consider hypothetical justifications for statutes when applying our three-part test. Therefore, while I agree that our three-part test arguably sets a different standard than the federal rational basis test, I cannot agree with McKnight that Wegan indicates that our three-part test requires us to consider only the actual justifications for a classification.
McKnight also points to our footnote in Wegan, which states that even if the statute was constitutional under the U.S. Constitution, the statute would still be unconstitutional under the Minnesota Constitution. This footnote indicates that the three-part test we used in Wegan may provide a different standard than the federal rational basis test; but, the footnote does not indicate that the basis for this different standard was refusing to provide a hypothetical justification for a challenged classification. I conclude that the lack of a discussion of a hypothetical justification for the challenged statute and our footnote in Wegan do not signify that we may consider only the actual justification for the statute.
In Russell, we also cited Nelson v. Peterson, 313 N.W.2d 580 (Minn.1981), to support the notion that we may look only *534to the actual justification for a challenged classification. Russell, 477 N.W.2d at 889. In my view, not only is this view of Nelson erroneous, our decision in Nelson may actually support the opposite conclusion. In Nelson, we considered a challenge to a statute that prohibited attorneys from becoming workers’ compensation judges if the attorney worked for the workers’ compensation division and represented employees. 313 N.W.2d at 580-81. Despite this prohibition, the statute allowed attorneys to become workers’ compensation judges if they worked for the division and represented the State. Id. at 581. In deciding that the statute violated equal protection rights, we considered hypothetical justifications for the statute, not just actual justifications provided by the Legislature. More specifically, we considered the State’s arguments about what the Legislature “may have legitimately believed,” and what the Legislature “could have believed” in enacting the statute at issue. Id. at 582-83 (internal quotation marks omitted). Moreover, we rejected a rationale for the law that we supplied. See id. at 582. Therefore, in Nelson, we considered hypothetical justifications for the statute raised by the parties and we supplied a hypothetical justification of our own. But after doing so, we simply dismissed these justifications as failing rational basis review.
The final case cited by Russell to support the proposition that we may not supply a hypothetical justification for a challenged classification is Thompson v. Estate of Petroff, 319 N.W.2d 400 (Minn.1982). In Thompson, the defendant and the deceased had an altercation, and the deceased attacked, raped, and threatened to kill the defendant. Id. at 401. The defendant killed the deceased during the attack, and was subsequently acquitted of a murder charge brought against her. Id. The defendant then brought an action against the deceased’s estate, seeking the recovery of damages resulting from the assault. Id. The district court dismissed the defendant’s action on the ground that the survival statute did not allow intentional tort claims to survive the perpetrator’s death, but did allow other types of claims to survive. Id. at 402. In Thompson, we applied the same three-part test utilized in Wegan and Nelson when we held that the omission of intentional torts from the survival statute violated equal protection. See id. at 404.
There are two reasons I believe that Thompson, like Wegan and Nelson, does not stand for the proposition that we may not consider hypothetical justifications for a classification. First, in Thompson, the decedent’s estate argued that the omission of intentional torts from the survival statute was based on the difficulty for the estate to defend against an accusation of intent without the perpetrator’s point of view. Id. The estate cited to no legislative source to support this argument.20 We considered the estate’s argument, even though it was not necessarily the actual justification for the classification drawn by the statute. See id. Second, in Thompson, we stated that we “regarded] the rational basis test as the same under both” the Minnesota and U.S. Constitutions. Id. at 406 n. 10. Therefore, in Thompson we implicitly disclaimed any sort of heightened standard in which a hypothetical justification would not suffice.
The foregoing analysis illustrates that the three cases cited in Russell do not provide adequate support for the notion *535that we may not supply a hypothetical justification for a statute when reviewing an equal protection claim under rational basis review. Therefore, I conclude that there is support for using hypothetical justifications in conducting rational basis review of legislative classifications.
Inconsistency in application
There are additional reasons why, when applying rational basis review to statutes facing constitutional challenges, we may provide hypothetical justifications for the statute. The three-part rational basis test we apply to equal protection challenges has a basis in our Uniformity Clause jurisprudence. See Miller Brewing Co., 284 N.W.2d at 356 & n. 3. Yet tax cases that implicate the Uniformity Clause and equal protection rights do not require that we consider only the actual justification for a challenged classification. In Council of Independent Tobacco Manufacturers of America v. State, a case addressing both a Uniformity Clause and an equal protection challenge, we stated that
in analyzing the purpose of the classifications created by the statute, we are not restricted to those purposes expressly stated by the legislature. Rather, any legitimate purpose can support the classifications created by the statute.
713 N.W.2d 300, 310 (Minn.2006). Similarly, when analyzing substantive due process claims under rational basis review, we do not require the actual justification be used. See Boutin v. LaFleur, 591 N.W.2d 711, 717-18 (Minn.1999) (citing a U.S. Supreme Court case considering a federal equal protection issue to outline the rational basis test we applied to a substantive due process claim).
Moreover, the justification for treating tax cases under the Uniformity Clause differently from other equal protection challenges is illogical. We have said that courts should be deferential to the Legislature in tax legislation because “taxation policy is peculiarly a legislative function, involving political give-and-take.” Minn. Automatic Merch. Council v. Salomone, 682 N.W.2d 557, 561-62 (Minn.2004) (citation omitted) (internal quotation marks omitted). Our explanation for the exception in tax cases could apply to almost all legislation enacted by the Legislature. Establishing the elements of a crime and the levels of punishment are also “peculiarly a legislative function,” just as is tax legislation. See State v. Olson, 325 N.W.2d 13, 17-18 (Minn.1982) (“The power to define the conduct which constitutes a criminal offense and to fix the punishment for such conduct is vested in the legislature.”).
Additionally, we have supplied a hypothetical justification for a classification in an equal protection challenge to a sentencing disparity case decided after Russell. In Behl, we considered an equal protection challenge to a sentencing disparity based on whether a juvenile was indicted, but not convicted, with first-degree murder. 564 N.W.2d at 568-69. We made no mention of our three-part test when applying the federal rational basis test. In applying the federal test, we stated that “[a]lthough the state failed at oral arguments to assert a rational basis upon which to base this different treatment,” the grand jury’s finding of probable cause supplied a rational basis. Id. at 569.
These inconsistencies, occurring in other circumstances involving constitutional challenges to statutes, support a conclusion that we may supply hypothetical justifications for classifications created by a challenged statute.
Separation of powers issues
There is another and more fundamental reason why I believe we must discontinue our refusal to allow courts to consider *536hypothetical justifications for classifications in equal protection cases. I conclude that requiring the Legislature to justify every piece of legislation, or else be subject to an equal protection challenge, offends the notion of separation of powers. To use the “actual” justification for the statute, we would need to look to the legislative record and parts of the statute itself. Requiring a justification for any classification appears to relegate the legislative and executive branches to a position inferior to the courts, requiring these branches to justify any classifications they create or else face being overruled. Rather, such a result violates well-established separation of powers principles. Legislative and executive branches have the power to work within the confines of the Minnesota and U.S. Constitutions and to make their own policy judgments, without having to justify non-suspect classifications to the courts.
Based on the foregoing multi-part analysis, I would conclude that in most equal protection cases we do not prohibit the consideration of hypothetical justifications for classifications when applying rational basis review.21 I see no reason for us to continue to refuse to utilize and assess hypothetical justifications for classifications facing rational basis review. Accordingly, to the extent that our three-part test has been distinguished from and viewed to establish a higher standard than the federal test based on a prohibition on considering hypothetical justifications, I conclude that such a distinction lacks merit.
D.
The preceding analysis outlines what I believe to be the proper approach to deciding an equal protection challenge under the Minnesota Constitution. I concede that my analysis does not reconcile every single case in our equal protection jurisprudence, but I believe such a total reconciliation of all of our disparate cases is not possible and, at this point, not necessary. As part of our obligation to the people of Minnesota, we must provide clear guidance on how individuals may vindicate their equal protection rights under the Minnesota Constitution. The test we apply may, if Cox is convicted of the charged offense, be the difference between whether Cox is branded as a felon or a gross misdemean-ant; whether the State retains the ability to put Cox in prison for a maximum term of five years or one year; and whether the State may fine Cox an additional $7,000. The stakes in this case, and in similar cases, are too high for us to make a mistake about the proper test to apply.
To summarize, the approach to equal protection cases (other than those involving a disparate impact on race, as in Russell ) that best reconciles our case law is as follows: First, we generally apply the similarly-situated test as a threshold test so that only claims that actually invoke equal protection concerns are reviewed. Second, we apply our three-part rational basis test rather than the federal version of the rational basis test. Our three-part test asks different questions than the federal rational basis test, and should be applied by the courts as it has been outlined in our recent opinions. Finally, we may assess hypothetical justifications in support of legislative classifications.
*537II.
Having articulated what I conclude to be the appropriate approach to be used by our court in equal protection challenges under the Minnesota Constitution, the next step is to apply this approach to the facts of the case before us. I begin, as the majority did, by applying the similarly-situated test. But, I conclude that the majority has incorrectly applied the similarly-situated threshold test. The majority incorrectly requires Cox to prove she is similarly situated with people convicted of theft by check, even though Cox should only be required to prove that she is similarly situated with people charged with theft by check. Therefore, the majority applies a standard that is too stringent.
A.
To succeed in her challenge, Cox must prove beyond a reasonable doubt that the Legislature violated the Minnesota Constitution by treating similarly-situated individuals differently without a rational basis. See Schober v. Comm’r of Revenue, 778 N.W.2d 289, 293 (Minn.2010) (stating constitutional violations must be proven beyond a reasonable doubt). At this point in the proceedings, it is unclear whether Cox could be convicted under either statute and Cox is innocent of all charges until proven guilty. Therefore, to meet the similarly-situated test, Cox must prove beyond a reasonable doubt that she is capable of being charged with theft by check. Prior cases involving challenges to punishment schemes required individuals convicted of one crime to prove they are similarly situated with individuals convicted of another crime. In Frazier, for example, the appellant had pleaded guilty to a crime with a more severe punishment and we concluded he was not similarly situated to individuals convicted of a crime with a less-severe punishment. 649 N.W.2d at 830, 839. Here, Cox is only charged with a crime that carries a more severe punishment, and we should therefore compare her with individuals charged with a crime that carries a less severe punishment. To require that Cox prove she could be convicted of theft by check would essentially require Cox to prove her own guilt of issuing a dishonored check before trial has commenced.
To prove Cox was similarly situated to those charged with theft by check, she must show that there was probable cause to support a charge of either theft by check or issuing a dishonored check. See Minn. R.Crim. P. 2.01, subd. 4. I will apply the differences between the elements of theft by check and issuing a dishonored check to the facts alleged in the complaint in order to decide whether Cox could have been charged with either crime. I conclude that Cox can show, beyond a reasonable doubt, that the State had probable cause to charge Cox under either statute based on the facts in the complaint and thus, Cox faces the punishment imposed under either statute.
Mens rea
The first difference between the two statutes is the mens rea requirements. The theft-by-check statute requires that a defendant issue a check with the “intent to defraud,” whereas the dishonored-check statute requires that a defendant issue a check that he or she “intends ... not be paid.” Minn.Stat. §§ 609.52, subd. 2(3), 609.535, subd. 2. The dishonored-check statute lists permissible inferences that are sufficient to show that the defendant intended the checks not be paid, whereas the theft-by-check statute does not list these inferences to determine intent to defraud. Intent “generally is proved circumstantially, by inference from words and acts of the actor both before and after the incident.” State v. Johnson, 616 *538N.W.2d 720, 726 (Minn.2000). The majority states that the facts alleged in the complaint do not support Cox’s as-applied challenge because the allegations would not support a conviction. But, as illustrated earlier, the majority should be analyzing whether there is probable cause to support a charge of theft by check, not whether the facts in the complaint support a conviction of theft by check.
In State ex rel. Hastings v. Bailey, the defendant challenged whether there was probable cause to support a charge of felonious intent to defraud by aid of a check. 263 Minn. 261, 262-63, 116 N.W.2d 548, 549-50 (1962). The facts in the record showed that the defendant agreed to purchase cattle for more money than he currently possessed. Id. at 263, 116 N.W.2d at 550. He told the seller that his brother would give him the remainder, and he gave the seller a check asking that the seller post-date it. Id. at 263, 116 N.W.2d at 550. When the seller cashed the check, there were insufficient funds. Id. at 263-64, 116 N.W.2d at 550. Additionally, the defendant said he would take the cattle to pasture, but the seller saw the cattle being resold. Id. at 263-64, 116 N.W.2d at 550. We held that probable cause was shown because
the evidence established that the complainant parted with property of the value of $4,100, for which he received a worthless check; and that the defendant falsely represented that the check would be honored on July 14. The evidence also established that instead of pasturing the cattle the defendant sold all or part of them and kept the proceeds.
Id. at 266, 116 N.W.2d at 552.
In the ease before us, the allegations facing Cox are sufficient to show probable cause for intent to defraud just as they were in Bailey. Cox allegedly received property in exchange for checks, and, by giving the check to the retail stores, she impliedly represented that the check would be paid when presented to the bank. See id. at 264-65, 116 N.W.2d at 550-51. Additionally, Cox allegedly went to the same store two days in a row to spend nearly $350, which could indicate a plan to exploit businesses that allowed her to purchase merchandise with worthless checks. This alleged two-day shopping spree took place at the end of the month, after she had already allegedly issued three worthless checks at other retail outlets. Finally, all five of the worthless checks identified in the complaint were given to retail stores in exchange for merchandise — none of the allegedly worthless checks were used to pay for bills.22 This consistency in only passing worthless checks to retail stores tends to show an intent to defraud, rather than simply a budgetary mistake. See City of St. Paul v. Greene, 238 Minn. 202, 204, 56 N.W.2d 423, 424-25 (1952) (stating that “a system of successive forgeries or of successive cheats or swindles of the same general nature” is admissible to show “that the defendant had been engaged in practicing like or similar cheats, as tending to prove a criminal intent” (citation omitted) (internal quotation marks omitted)). All of these allegations show that the State had probable cause to charge Cox with either offense.
Obtaining property or services
The next difference between the statutes is the requirement in the theft-by-check statute that the defendant obtains possession, custody, or title to property of another or obtains services from another. This *539requirement in the theft-by-eheck statute is not contained in the dishonored-check statute. In this case, the complaint alleges that Cox did receive possession, custody, or title to property of another because she allegedly used her checks at retail stores “as payment for merchandise.” Because the complaint does allege Cox obtained possession, custody, or title to property of another, she could have been charged under either statute.
Value of property or services obtained
Finally, the proof of dollar value is different under the two statutes. The theft-by-check statute bases punishment on the value of the goods or services obtained, whereas the dishonored-check statute bases punishment on the dollar amount of the check. I conclude that this element is inconsequential when applied to the facts of this case. Issuing dishonored checks “as payment for merchandise” at retail stores23 will presumably provide the issuer ■with goods of a value equal to the amount on the check. See Minn.Stat. § 609.52, subd 1(3) (2010) (defining “value” for the theft-by-check statute as “the retail market value at the time of the theft”).
Case law relied on by the majority is distinguishable
The majority compares this case to Frazier and Dietz in concluding that Cox is not similarly situated to those individuals convicted of theft by check. See State v. Frazier, 649 N.W.2d 828 (Minn.2002); State v. Dietz, 264 Minn. 551, 119 N.W.2d 833 (1963). I conclude that these comparisons are inapposite.
In Frazier, we considered the defendant’s as-applied challenge to the statute, and we stated that “the critical inquiry is whether the elements” of the two statutes were “the same or essentially similar.” 649 N.W.2d at 837. This quoted passage appears to reinforce the majority’s application of the similarly-situated test that looks to the general elements of each crime. Frazier, however, is distinguishable because in that case it was impossible to convict the defendant under both statutes. In Frazier, the punishment disparity was based on the number of predicate criminal acts committed by a defendant. Id. at 839. Minnesota law criminalized committing a crime for the benefit of a gang, and required only one predicate criminal act. See Frazier, 649 N.W.2d at 839 (citing Minn.Stat. § 609.229 (2000)). The RICO statute, on the other hand, criminalized a pattern of criminal activity, and required the State to prove at least three predicate criminal acts. Id. In Frazier, the defendant was convicted under section 609.229 based on his commission of one predicate criminal act. 649 N.W.2d at 839. Therefore, we stated that the defendant “is not similarly situated to an individual convicted under RICO, whose conviction must be based on his participation in at least three criminal acts.” Id. Here, unlike in Frazier, the allegations in the complaint are sufficient to show that Cox could have been charged under either statute, and therefore Cox is similarly situated with individuals charged under either statute.
In Dietz, the punishment disparity challenged by the defendants was based on the dollar value of the property stolen. 264 Minn. at 558-59, 119 N.W.2d at 837-38. In Dietz, the punishment for stealing property worth less than $25 from an automobile at night was greater than the punishment for stealing property worth between $25 and $100 from an automobile at night. Id. at 554, 119 N.W.2d at 835. The defendants in Dietz were charged with stealing property worth less than $25 from an auto*540mobile at night. Id. at 552, 119 N.W.2d at 834. Under the statutory scheme, the defendants’ actions could be subject only to the punishment attached to one level of theft, but not any of the others. Therefore, the defendants in Dietz could only be charged with the act of stealing property worth less than $25, and could not be charged with the act of stealing property worth between $25 and $100. Therefore, the defendants were not similarly situated with others who were charged with stealing property worth between $25 and $100. In this case, unlike in Dietz, Cox’s alleged acts could lead to charges under either statute.
Based on the preceding analysis, I would conclude Cox is similarly situated with individuals charged with either theft by check or with issuance of a dishonored check. Therefore, I would hold that Cox meets the similarly-situated threshold test, and I would proceed to apply rational basis review to her equal protection challenge.
B.
Having established that Cox is similarly situated to those individuals charged with either theft by check or issuing a dishonored check, I now apply our rational basis test. The three-part rational basis test we apply under the Minnesota Constitution requires the following:
(1) The distinctions which separate those included within the classification from those excluded must not be manifestly arbitrary or fanciful but must be genuine and substantial, thereby providing a natural and reasonable basis to justify legislation adapted to peculiar conditions and needs; (2) the classification must be genuine or relevant to the purpose of the law; that is there must be an evident connection between the distinctive needs peculiar to the class and the prescribed remedy; and (3) the purpose of the statute must be one that the state can legitimately attempt to achieve.
Russell, 477 N.W.2d at 888. I would conclude that the punishment disparity between the dishonored-check statute and the theft-by-check statute meets part 1 of our three-part test, but cannot meet part 2 of that test. Because part 2 of our test cannot be met, there is no need to address part 3.
Part 1
The first question is whether the distinction between individuals charged with committing theft by check and individuals charged with issuing a dishonored check is “manifestly arbitrary or fanciful” or whether there is a “genuine and substantial” distinction between the two groups. As earlier noted, there are three differences between the two offenses — mens rea, proof of obtaining property or services, and the method of determining the punishment level. These distinctions between the two groups are not arbitrary or fanciful, but provide a “natural and reasonable basis to justify” differing treatment of the groups. The Legislature is certainly free to treat a more-culpable mens rea differently than a less-culpable mens rea. Also, the Legislature could reasonably believe that people who use false checks to obtain property or services deserve different treatment than those who use false checks and receive nothing additional in return.24 Therefore, I conclude the stat*541utes meet the first part of our three-part rational basis test.
Part 2
The next question is whether the classification is “genuine or relevant to the purpose of the law.” The purposes of the criminal code are to prevent crime with deterrence, to rehabilitate the convicted, to confine the convicted when required, and
to protect the individual against the misuse of the criminal law by fairly defining the acts and omissions prohibited, authorizing sentences reasonably related to the conduct and character of the convicted person, and prescribing fair and reasonable postconviction procedures.
MinmStat. § 609.01 (2010) (emphasis added). We have held that issuing a dishonored check is a lesser-included offense of theft by check. State v. Roden, 384 N.W.2d 456, 458 (Minn.1986). By providing a greater punishment for the lesser-included offense, the Legislature renders the classification irrelevant to the purpose of providing sentences that are reasonably related to the conduct charged. This punishment scheme turns the notion of lesser-included and greater-included offenses on its head and could lead to arbitrary charging decisions by the State. Moreover, the Legislature’s purpose in creating separate crimes for greater and lesser offenses is not served. Additionally, the Legislature is failing to deter more-culpable criminal acts by removing any incentive to not commit the greater offense instead of the lesser offense. The State did not provide adequate justification for this result, and I cannot conceive of any rational hypothetical reason that would prompt the Legislature to desire this outcome. A mistake or inadvertence, the only basis for the sentencing disparity advanced by the State, is not a reasonable basis for the classification. Therefore, I conclude that the classification is irrelevant to the purpose of the law, and that therefore, the punishment disparity between the theft-by-check statute and the dishonored-check statute as applied to Cox is a violation of equal protection.
Part 3
Because I conclude that the classification is not relevant to the purpose of the law, I do not need to address part 3 of our three-part rational basis test — whether the purpose of the law is legitimate.
Additional Support
My conclusion that Cox has asserted a valid equal protection challenge to the dishonored-checks/theft-by-check statutes is supported by our decision in State v. Garcia, 683 N.W.2d 294 (Minn.2004). In Garcia, as previously noted, a statute provided different rules for whether an individual would receive jail credit for time served based on the individual’s certification as an adult, or certification as an extended jurisdiction juvenile (“EJJ”). Id. at 298. EJJs who violated probation would not receive certain jail credit that would be given to juveniles certified as adults who subsequently violated their probation. Id. at 297-98. We held that this disparity was not rational. Id. at 301. The reason for the disparity provided by the State was that a greater punishment was given to EJJs because EJJs are more amenable to treatment, and the threat of a longer prison sentence would be helpful. Id. at 299-300. We concluded that there was no rational basis for this distinction, because the purpose of probation is the same for all of juveniles on probation. Id. at 300. Therefore, we stated that “it is not rational, under these circumstances, to punish juveniles designated as EJJs more severely than juveniles certified as adults who engage in the same conduct.” Id.
*542Here, I also conclude there is no rational reason to punish a person who violates the dishonored-check statute more harshly than a person who violates the theft-by-check statute. The facts of this case present a stronger case for an equal protection challenge than the facts in Garcia. In Garcia, we held that the purpose of probation for both groups was to incentivize individuals to abide by the rules, and there was no reason to provide one group with a greater incentive. But here, there is a reason to provide a harsher penalty for one group because one crime is the lesser-included offense of the other. But, the Legislature did the complete opposite.
For all the foregoing reasons, I conclude that the sentencing disparity between the theft-by-check statute, Minn.Stat. § 609.52, subd. 3(4), and the dishonored-check statute, Minn.Stat. § 609.535, subd. 2a(a)(l), is a violation of equal protection. Therefore, I would answer the certified question in the affirmative, and reverse the court of appeals.
. We stated in State v. Roden, 384 N.W.2d 456, 458 (Minn.1986), that issuing a dishon*527ored check is a lesser-included offense of theft by check.
. In some older cases, we applied a different test that appeared to combine aspects of the similarly-situated test and the post-1979 three-part test. This older test stated that equal protection "requires that a legislative classification apply uniformly to all those similarly situated; that the distinctions separating those who are included within the classification from those who are excluded are not arbitrary or capricious, but instead are real and substantial; and that the classification is consonant with a lawful purpose.” Kossak v. Stalling, 277 N.W.2d 30, 34 (Minn.1979). But, this test has not been used by our court since the early 1980s, except that a slightly different version has been used specifically for equal protection challenges to classifications created by workers’ compensation statutes. In Gluba ex rel. Gluba v. Bitzan & Ohren Masonry, we articulated the test for workers’ compensation statutes as follows: " 'To survive [an equal protection] challenge, a [workers’ compensation] classification must apply uniformly to all those similarly situated; be necessitated by genuine and substantial distinctions between the two groups; and effectuate the purpose of the law.’ ” 735 N.W.2d 713, 721 (Minn.2007) (quoting Bituminous Cas. Corp. v. Swanson, 341 N.W.2d 285, 287 (Minn.1983) (alterations in original)). We stated that this test was different from the three-part test normally applied to equal protection challenges. Id. at 722-23.
. This reconciliation is only necessary for equal protection challenges brought under the Minnesota Constitution. If the federal standard for analyzing equal protection challenges needs clarification, the federal courts should work on the clarification.
. Federal case law supports this approach. See, e.g., Keevan v. Smith, 100 F.3d 644, 648 (8th Cir.1996) (“[T]he initial inquiry in any equal protection claim is whether the plaintiff has established that she was treated differently than others who are similarly situated to her.”); Klinger v. Dep’t of Corr., 31 F.3d 727, 731 (8th Cir.1994) ("Absent a threshold showing that she is similarly situated to those who allegedly receive favorable treatment, the plaintiff does not have a viable equal protection claim.”).
. The majority's opinion in this case reinforces this point. The similarly-situated test is used by the majority to deny the equal pro tection claim before the majority identified a level of review or applied that level of review.
. See, e.g., Greene v. Comm’r of Minn. Dep’t of Human Servs., 755 N.W.2d 713, 729-30 (Minn.2008); State v. Barnes, 713 N.W.2d 325, 332 (Minn.2006); State v. Benniefield, 678 N.W.2d 42, 46 (Minn.2004); Westling v. Cnty. of Mille Lacs, 581 N.W.2d 815, 820 (Minn.1998); State v. Russell, 477 N.W.2d 886, 888 (Minn.1991); McGuire v. C & L Rest. Inc., 346 N.W.2d 605, 612 (Minn.1984); Thompson v. Estate of Petroff, 319 N.W.2d 400, 404 (Minn.1982); Wegan v. Vill. of Lexington, 309 N.W.2d 273, 280 (Minn.1981); Miller Brewing Co. v. State, 284 N.W.2d 353, 356 (Minn.1979).
. It is not obvious from our opinion in Behl whether we were applying the federal rational basis test to an equal protection challenge under the U.S. Constitution or the Minnesota Constitution. The appellant in Behl did not cite to either constitution in his equal protection argument, but he did cite to the constitutional law portion of Dunnell Minnesota Digest, indicating a challenge brought under the Minnesota Constitution.
. Cox only argues for us to apply the rational basis test. Therefore, this dissent is focused exclusively on cases that apply a rational basis test, and takes no position on when we should apply strict or intermediate scrutiny.
. Specifically, we cited to McKnight’s discussion of the following cases: Wegan v. Vil*533lage of Lexington, 309 N.W.2d 273 (Minn.1981), Nelson v. Peterson, 313 N.W.2d 580 (Minn.1981), and Thompson v. Estate of Petroff, 319 N.W.2d 400 (Minn.1982). Russell, 477 N.W.2d at 889.
. Respondents cited to legislative history only to show that the Legislature declined to add intentional torts to the survival statute. When discussing the rational basis for the Legislature's decision, no legislative sources were cited.
. In Russell, we dealt with a statute that had a disparate impact based on race. We concluded that it was "particularly appropriate that we apply our stricter standard of rational basis review in a case such as this where the challenged classification appears to impose a substantially disproportionate burden on the very class of persons whose history inspired the principles of equal protection.” Russell, 477 N.W.2d at 889. I would leave our Russell opinion undisturbed as applied to cases alleging a disparate impact based on race.
. The complaint alleges Cox passed worthless checks numbered 1151, 1159, 1163, 1164, and 1165 on the same account. Therefore, presumably, there were at least eleven other checks issued by Cox during December that were paid or not reported.
. Cox allegedly issued dishonored checks at Benson Liquor Store, Darold’s Super Valu, Kelly’s Convenience Store, and Running's Farm & Fleet.
. Although I concluded that the allegations against Cox provide probable cause to meet the requirements of both statutes, this conclusion does not mean that the classification drawn by the statutes is manifestly arbitrary or fanciful. The distinction drawn by the statutes is a genuine and substantial difference relevant to a defendant's culpability. The ability to charge Cox with either offense *541does not mean that the distinction between the offenses is arbitrary.