dissenting:
I regret I cannot join in the court’s opinion. There are two impediments. The first is the court’s conclusion that an inventory search of the automobile was justified over explicit objection. The key analytical problem is the court’s assertion that “it was necessary to secure the vehicle and its contents.” It is, of course, fairly settled that when officers remove a defendant from his car and arrest him, they can take steps to secure the car. I would assume we would not uphold the arrest of a shoplifter in a store and allow the police to inquire where his or her car is parked and then find the car and determine it “necessary to secure the vehicle and its contents.” In this case, the defendant had parked his car in a lawful place and left it there to go inside to do business. It was there that he was arrested. I believe it goes well beyond any established Supreme Court precedent and beyond logic to ratify an “inventory” search based on the assertion that securing this vehicle under these circumstances was either necessary or permissible. The suggestion of some police duty to the car here argues too much. It was not unlawfully parked. It was not a traffic hazard. They had not stopped him while driving or otherwise removed him from his car. His arrest was no more associated with this car than the arrest of a shoplifter would be associated with his or her car. The facts of this case make what one often suspects is pretext, patent.
We previously invalidated an inventory search, under similar circumstances, in United States v. Pappas, 735 F.2d 1232 (10th Cir.1984). There the defendant was indicted for possession of an illegal weapon found in his car. The evidence was conflicting as to whether the defendant entered his car prior to the arrest. The arresting officer testified that the defendant had entered the vehicle. At a minimum defendant was headed toward the car and about to enter it. In that case there was clearly a closer nexus between the defendant and the inventoried vehicle. We held that while the protective search of the compartment in defendant’s car was reasonable, that initial search did not “extend to the inventory search of defendant’s car.” Id. at 1234. We distinguished South Dakota v. Opperman, 428 U.S. 364, 96 S.Ct. 3092, 49 L.Ed.2d 1000 (1976):
Opperman cannot be used to justify the automatic inventory of every car upon the arrest of its owner. The justifications for the rule are too carefully crafted for this to be the intent. In this case the car was parked on private property and there was no need for the impound and inventory search.
*720Id. at 1234. There was likewise no need for the impound and inventory here.
I am also unable to agree that the willful introduction of inadmissible evidence over specific court instruction is harmless. While we often indulge the tenuous notion that a curative instruction in fact removes taint and often suggest that after the prosecutor has battled tooth-and-nail to get inadmissible evidence before the jury that it didn’t influence the verdict, I don’t think that is what we are really saying. Basically I believe we are just acknowledging that some mistakes have to be borne by defendants. In this case I believe we should not extend that principle which puts the burden of risk on the defendant where that risk was willfully created. While there is substantial evidence of guilt, the prosecutor knew that and still felt very strongly that this inadmissible evidence was important in persuading the jury. The court found that the prosecutor’s conduct was willful. The court, by its cautionary instruction to the jury, puts the lie to any suggestion that it admitted this evidence as proper. I simply am unable to say that something so important to the prosecutor that she three times brought it before the jury, twice risking contempt, is harmless beyond a reasonable doubt. I just cannot say that a jury would not be influenced by this evidence in what otherwise appears to be a strong case.
I would grant a new trial.