(dissenting).
I respectfully dissent from the majority’s approval of the admission of Spreigl evidence in this instance. The majority seems to believe that, while this defendant’s previous offenses do not rise to the level of similarity required of so-called signature crimes, they are nonetheless simply “similar enough.” This strikes me as reminiscent of the old adage “close enough for government work” and is certainly not an adequate basis for the admission of damaging, highly prejudicial evidence.
Our prior cases require, not request or encourage, the articulation of the nature of the supposed similarity between the instant and prior offenses. Spreigl evidence is admissible for the legitimate purpose of establishing motive, intent, absence of mistake, identity, or common purpose or plan. State v. DeWald, 464 N.W.2d 500, 502-03 (Minn.1991). This court has held that so-called prior crime evidence may be admitted if it is similar to the charged offense in either time, place, or modus operandi. State v. Filippi, 335 N.W.2d 739, 743 (Minn.1983). However, as one of the majority’s quotations points out, that the prior crime is simply of the same generic type as the instant offense is insufficient to allow its admission as Spreigl evidence. State v. Cogshell, 538 N.W.2d 120, 123-24 (Minn.1995).
Quoting Cogshell, the majority implies, but does not state, that the prior crimes admitted in this ease are similar enough in terms of their modus operandi to prove identity. Ante, at 363. Yet, the only similarity between the crimes is that they were robbery/assaults — that is, of the same generic type. The prior crimes occurred three and a half years previously, involved robbery/assaults, which occurred at night, in the ATM vestibule of a Minneapolis bank, and involved victims who had just withdrawn money from the cash machines; in each of them, appellant was the getaway driver but was not otherwise directly involved. In each, the victims were beaten by their assailants, who, as noted, did not include the appellant. In each, the weapon was a bat or tire iron.
The facts of the crime at issue here could hardly be more different and still bring it within the same section of the Minnesota criminal code. The robbery/assaults at issue here did not involve individuals who had just left a cash machine, but rather were directed at individuals either walking or driving through the streets of St. Paul. Here, in one instance the assailants approached the victim in an automobile and discharged firearms from inside the automobile, and in the other, the victim himself was inside an automobile. No bats or tire irons were used and the appellant is charged with being the main assailant, not merely the driver. These facts are very different and appear to violate the requirement of Cogshell that the acts not only comprise a violation of the same statute, but that they be factually similar. One must ask: where is the identity at which the majority hints? Where is the modus operandi?
While trial courts certainly merit the considerable discretion invested them to make close evidentiary rulings, they are not infallible and I believe the admission of this prior crimes evidence was an abuse of discretion. Trial errors of this type are subject to a harmless error analysis, State v. Howard, 324 N.W.2d 216, 222-23 (Minn.1982), cert. denied, 459 U.S. 1172, 103 S.Ct. 818, 74 L.Ed.2d 1016 (1983), and I also believe, reviewing the weak nature of the state’s ease, that this error was not harmless. I would, therefore, reverse the trial court and remand the matter for a new trial.
I have recorded my dissatisfaction with the use of Spreigl evidence on several occasions. See, e.g., State v. Cogshell, 538 N.W.2d 120, 124 (Minn.1995); State v. Walsh, 495 N.W.2d 602, 607 (Minn.1993); State v. Berry, 484 N.W.2d 14, 19 (Minn.1992) (dissenting opinions of Gardebring, J.) The Spreigl decision, I believe, all too frequently allows the introduction of prejudicial evidence precisely when it most ought to be excluded, namely, when the state’s ease is weak. See Cogshell, *366538 N.W.2d at 125 (Gardebring, J., dissenting).
The evidence described above was not probative of the identification of the perpetrator of the crime and should not have been admitted. Its admission allowed the jury to bolster its decision to convict with the assumption that, even if this defendant did not commit this particular crime, he has committed violent crimes in the past and therefore is likely to do so again, so it is best just to use this opportunity to put him behind bars. To affirm its admission with scant analysis to illuminate how and why this evidence rises to the level of similarity required by our prior decisions serves only to perpetuate the problems spawned by Spreigl.