(dissenting) — If the majority is willing to convict DWI defendants solely on the "testimony” of a machine, it should provide a method to ensure the machine operates correctly. Yet, notwithstanding the wealth of evidence present in the record relating to machine errors, unauthorized machine contacts, and overall DataMaster accuracy, the majority places almost no burden on the State to prove the BAC DataMaster Verifier (DataMaster) machines do, in fact, yield correct results.
The majority confuses a "valid” test result with one that is accurate and reliable. The majority presumes if the Data-Master spits out a number, that number is correct. Only by ignoring the record can the majority make this gigantic leap of faith. As forensics expert Dr. Richard Jensen explains:
[A] test may be "valid” ie; properly performed and completed, based upon [the State Toxicologist’s] subjective criteria, but still be inaccurate and unreliable. One must necessarily look to the evidence extrinsic to the breath test document to consider the latter question, evidence which is not being preserved under the program currently in force in Washington.
Declaration of Richard E. Jensen, Ph.D., in Response to State’s "Clarifying” Affidavits, at 7-8.
Rather than requiring the necessary extrinsic evidence, the majority instead accepts a "self-certifying” breath test ticket as the sole proof of reliability and accuracy, and denies the defense any means of challenging this automated justice dispenser. I reject this approach because it is patently unfair and because it violates the due process clauses of the state and federal constitutions.
At issue is whether a criminal defendant’s constitutional due process right to a fair trial is violated by the State’s failure to maintain repair and maintenance records on the machines used to measure breath alcohol. In Washington, destruction of evidence cases have been governed by a 2-part test that weighs the likelihood the destroyed evidence would have been exculpatory against the ability of the prosecution to have preserved the evidence. State v. Vaster, 99 Wn.2d 44, *492659 P.2d 528 (1983); State v. Wright, 87 Wn.2d 783, 557 P.2d 1 (1976).
The majority rejects this well-established test in favor of the federal analysis adopted in Arizona v. Youngblood, 488 U.S. 51, 102 L. Ed. 2d 281, 109 S. Ct. 333 (1988), and California v. Trombetta, 467 U.S. 479, 81 L. Ed. 2d 413, 104 S. Ct. 2528 (1984), and concludes the State’s failure to preserve the evidence does not violate Defendants’ due process rights. I disagree with this result because the majority has incorrectly analyzed the state constitutional issue under State v. Gunwall, 106 Wn.2d 54, 720 P.2d 808, 76 A.L.R.4th 517 (1986), and because the Youngblood and Trombetta standard does not provide adequate due process protection as interpreted by this court.
I
The outcome of these cases is controlled by the well-established "reasonable balance” test of Vaster and Wright. The majority rejects , this analysis because it erroneously concludes these cases are no longer valid in light of Young-blood and Trombetta. The majority contends Vaster and Wright are not state constitutional cases because they rely on federal law. I disagree. Vaster and Wright are still good law.
Neither Vaster nor Wright relies on federal law or cites specifically to the federal constitution. Rather, the cases evaluate both federal cases and cases from other jurisdictions, and, from these cases, form an independent due process test for destruction of evidence cases. Citations to federal cases are not determinative on the issue whether a case is decided on state or federal constitutional grounds. This court often looks to federal cases as persuasive authority on state constitutional issues, even though these cases are not controlling. See Rozner v. Bellevue, 116 Wn.2d 342, 351, 804 P.2d 24 (1991) (although not controlling, federal due process decisions are afforded great weight).
Vaster in particular makes it clear this court felt unconstrained by federal law, and instead developed a test which provided more protection for this state’s citizens than does *493the federal constitution. Discussing its earlier decision in Wright, the Vaster court explained:
Unlike other jurisdictions which have focused on the good or bad faith of the prosecution in destroying the evidence, we preferred to adopt the second Agurs standard and look to the reasonable possibility that the destroyed evidence was material and favorable to the defendant. See Wright, at 792.
(Footnote omitted.) Vaster, 99 Wn.2d at 50. As an example of another jurisdiction’s analysis rejected by the court, the Vaster court cited to United States v. Augenblick, 393 U.S. 348, 21 L. Ed. 2d 537, 89 S. Ct. 528 (1969), a Supreme Court decision adopting a good faith/bad faith approach when evaluating the suppression of certain evidence.
Because Vaster and Wright are still good law, it is unnecessary for the majority to apply the Gunwall analysis. The purpose of Gunwall is to determine whether the Washington Constitution provides a level of protection greater than the federal constitution in a given case. Gunwall, 106 Wn.2d at 61. Surely we are not, however, required to revisit each of our state constitutional cases decided prior to Gunwall. Here, the constitutional question has already been answered. Vaster and Wright articulate a separate test from the federal cases they cite and they provide greater protection than the federal constitution. Most important, however, is that the Vaster and Wright "reasonable balance” test is well established, workable, and familiar to trial courts in this state. There is no reason to abandon this test in favor of Youngblood’s less protective "bad faith” standard.
Even if an independent state constitutional analysis is reached under Gunwall, the same result would be attained. In State v. Ortiz, 119 Wn.2d 294, 831 P.2d 1060 (1992), the dissent argued, under Gunwall, that the Vaster test is the proper standard to apply in destruction of evidence cases. Ortiz, 119 Wn.2d at 316-20 (Johnson, J., dissenting). I adhere to those arguments and raise only a few additional points regarding the majority’s analysis.
The Defendants here rely on Vaster and Wright as preexisting state law in support of an independent state con*494stitutional analysis. The majority again rejects these cases, however, because they do not expressly include an analysis of the state constitution. This approach misconstrues the analysis underlying GunwalVs fourth factor. "Factor four requires us to examine preexisting state law to determine what kind of protection this state has historically accorded the subject at issue”. State v. Young, 123 Wn.2d 173, 179, 867 P.2d 593 (1994) (citing Gunwall, 106 Wn.2d at 61-62). Gunwall and its progeny do not limit our inquiry under this factor to only those cases engaging in a lengthy state constitutional analysis. Gunwall itself is significantly more broad. It directs courts to examine "[previously established bodies of state law, including statutory law . . .”. (Italics mine.) Gunwall, 106 Wn.2d at 61. The purpose of this factor is to permit courts to evaluate the importance of the particular right being asserted in light of this state’s unique legislative and common law background.1
Here, the majority fails to recognize that cases decided prior to Gunwall are not likely to engage in the same highly structured state constitutional analysis as would postGunwall cases. This does not mean these cases do not reflect particular state interests. Vaster’s "reasonable balance” test embodies this State’s strong interest in preserving a crimi*495nal defendant’s constitutional due process right to a fair trial.
Under Vaster’s "reasonable balance” test, a court evaluates destruction of evidence cases by weighing the exculpatory potential of a piece of lost evidence against the State’s ability to preserve that evidence. Applying the Vaster test in these cases, I would find Defendants’ due process rights were violated by the State’s failure to preserve the maintenance and repair records.
Vaster’s first prong requires us to determine if there was a reasonable probability the missing evidence would have been exculpatory. The trial court in State v. Wittenbarger, Snohomish Cy. cause 91-2-01742-3, specifically found a reasonable probability exists that the records of voltage values, calibration factors, machine malfunctions, complaints by operators, and the records of repairs affect a person’s ability to defend against a DWI charge. The trial court further found a repair record is an important document in the evaluation of the accuracy of a machine.
The majority suggests the "repair and maintenance records would merely be used by the defense to discredit the general reliability of the DataMaster results”. (Italics mine.) Majority, at 476. Precisely! When the prosecution’s case depends upon a single number, the exculpatory value of these records lies in their ability to cast reasonable doubt upon the accuracy of the prosecution’s evidence. The expert testimony establishes the DataMaster is capable of completing a test, and, yet the result may be inaccurate and unreliable. That these records have been used successfully in the past to raise this doubt is ample proof that there is, at least, a "reasonable possibility” the maintenance and repair records would have proven exculpatory.
Finding a reasonable possibility the evidence would have been exculpatory, Vaster next requires us to weigh that possibility against the State’s ability to have preserved the evidence. This factor overwhelmingly supports finding a due process violation in these cases. The State had every ability to preserve the maintenance and repair records with little if *496any added effort. The State did, in fact, maintain these records for many years, and it took an overt policy change to prevent the State from continuing to maintain these records.
In light of the strong possibility the repair and maintenance records would have proved exculpatory, and the relative ease with which the State could have maintained these records, I would hold the State’s failure to do so violates the Defendants’ due process rights under article 1, section 3 of the Washington State Constitution.
II
The majority erroneously rejects the Defendants’ state constitutional analysis, and thus analyzes these cases under federal constitutional law. Because Vaster and Wright are still good law and provide more protection than the federal constitution, it is unnecessary to engage in a federal constitutional analysis. I only do so here because I believe the majority has incorrectly applied the federal due process test in these cases.
"Under the Due Process Clause of the Fourteenth Amendment, criminal prosecutions must comport with prevailing notions of fundamental fairness”. Trombetta, 467 U.S. at 485. This standard of fairness requires that the State afford criminal defendants a meaningful opportunity to present a complete defense. Trombetta, 467 U.S. at 485. The State has a duty to disclose material exculpatory evidence. United States v. Agurs, 427 U.S. 97, 49 L. Ed. 2d 342, 96 S. Ct. 2392 (1976); Brady v. Maryland, 373 U.S. 83, 10 L. Ed. 2d 215, 83 S. Ct. 1194 (1963). Alternately, the State violates a criminal defendant’s due process rights when it fails to preserve material exculpatory evidence, regardless of whether the State acted in good or bad faith. Arizona v. Youngblood, 488 U.S. at 57.
In recent years, two United States Supreme Court cases have shaped the test to determine whether the government’s failure to preserve the evidence violates a defendant’s right to due process. In Trombetta, the Court held the government violates the defendant’s right to due process if the evidence *497possessed "exculpatory value that was apparent before the evidence was destroyed, and be of such a nature that the defendant would be unable to obtain comparable evidence by other reasonably available means”. Trombetta, 467 U.S. at 489.
Later, in Youngblood, the Court added the caveat that when the evidence sought by the defense is not "material exculpatory” evidence, but is nonetheless potentially useful to the defense, the defendant must demonstrate the police acted in bad faith in failing to preserve the evidence. Young-blood, 488 U.S. at 57-58.
Although the majority articulates substantially the same federal test, it dances around these due process requirements, and ignores the central premise that the State must afford criminal defendants a meaningful opportunity to present a complete defense. Trombetta, 467 U.S. at 485.1 would find the State’s failure to generate maintenance and repair records violates the requirements of both Youngblood and Trombetta.
A
First, under Youngblood it would be difficult to imagine a clearer case of bad faith on the part of the government. The records in the proceedings below are replete with testimony from state officials who admitted they understood the importance of this testimony to the defense bar. Dr. Logan, the State Toxicologist, testified he was aware that defense attorneys were successfully challenging breath tests using repair and maintenance history of the DataMaster. He further indicated he viewed his role as helping prosecutors perform better in court against a vociferous defense bar.
Yet, by failing to maintain repair and maintenance records, the State systematically and intentionally established a policy that deprives the defense of a means of challenging the DataMaster test results. Ms. Carol Murren, the defense expert, testified the maintenance and repair records provide significant information about problems with the machines, and are necessary to provide an independent *498assessment of the accuracy of the tests. Defense expert Dr. Richard Jensen likewise explained that the repair and maintenance records, such as the telephone complaint records and the technician’s diagnosis of observed malfunctions, "contain essential information in considering the accuracy and reliability of a breath test . . . [because] they contain the actual, immediate observations of the malfunction by the person viewing it at the time it happens”. Declaration of Richard E. Jensen, Ph.D., at 5. Moreover, the State prevents anyone not employed by the Washington State Patrol from either inspecting the machines or becoming certified to operate them. In contrast, the State Toxicologist’s office does certify people outside of law enforcement to test and analyze blood samples.
The State’s actions are an indefensible attempt to streamline DWI prosecutions by eliminating the defenses’ tools. I would agree with the lower courts, none of which had any trouble finding bad faith on these facts:
In the present case the facts indicate that the only logical conclusion is that the Washington State Patrol has simply decided that it will not be a part of any effort to tarnish the reputation of its BAC machine. It has done so by systematically reducing the number of complaint and repair/adjustment records maintained on the BAC machine, as well as not allowing anybody outside of law enforcement to become a qualified/ certified operator and thus ... be able to run a subsequent comparison test as contemplated by RCW 46.61.506(5).
Cascade Dist. Court Dec. 6094185, at 12 (Mar. 11, 1991); Opening Br. of Pet’r (Wittenbarger) attach. A.
Likewise, in State v. Matthews, the court found:
The systematic elimination or failure to preserve certain records amounts to destruction of 'material’ evidence and was done in 'bad faith’, i.e. with knowledge of the potential exculpatory value of the evidence.
King Cy. Dist. Court K09034 Order Suppressing Def.’s Breath Test Results, at 2 (Feb. 22, 1993); Br. of Resp’t (Matthews) app. B.
The fact the majority finds certain legitimate reasons for the change in policy does not change this result. The repair *499and maintenance records sought by the defense are easy to maintain. They have always been maintained in the past. Most importantly, they are of undisputed value to the defense and to the truth-finding function of the court. Under these circumstances, the burden upon the State should be significant to demonstrate why the shift in policy was not an act of bad faith. Instead, the majority accepts wholesale the State’s thinly disguised and unsupported rationales, none of which justifies such a dramatic shift in longstanding policies. Majority, at 477-79.
B
The State’s failure to generate maintenance and repair records also violates the requirements of Trombetta. Under Trombetta, the failure to preserve material evidence violates due process unless the defendant is able to obtain comparable evidence by other reasonably available means. Trombetta, 467 U.S. at 489. Neither prong of this requirement is satisfied in this case.
The majority contends the Defendants in these cases have ample other means of presenting an adequate defense. This is simply not true. Prior to the 1992 protocol modifications, the defense bar relied heavily on the maintenance and repair records to attack the credibility of DataMaster breath tests. In some of these cases, the defense raised a reasonable doubt in the minds of the jurors as to the reliability of the breath test results.
Here, the majority eliminates this evidence. The defendants now are limited to cross-examining DataMaster operators regarding operator error, to examining experts regarding the DataMaster machines and infrared spectroscopy, and to conducting additional independent breath or blood tests obtained under RCW 46.61.506(5). Majority, at 476. However, with the exception of the expert testimony, none of the evidence described by the majority attacks the reliability of the DataMaster machinery — the touchstone of the defense. A person charged with a DWI can be found guilty if he or she has a breath alcohol test reading of .10 grams or higher. *500Former RCW 46.61.502(1). When the State’s entire case revolves around a DataMaster test result, a defendant must, by necessity, challenge factors underlying the accuracy of the test. Yet the majority finds no defects in a policy that permits the State to rely on potentially inaccurate results and then systematically eliminate the defense’s ability to challenge these results.
Even expert testimony regarding the DataMaster machines is inadequate to establish possible inaccuracies. In United States v. Cooper, 983 F.2d 928 (9th Cir. 1993), two defendants were charged with various offenses related to the manufacture of methamphetamine. Prior to trial, the government destroyed all of the laboratory equipment it seized it connection with the prosecution. The defendants argued the destroyed equipment was critical to their defense, alleging the laboratory equipment was physically incapable of producing methamphetamine, and the destroyed equipment would have established their innocence. Relying on Trombetta and Youngblood, the District Court dismissed the prosecution.
On appeal, the government challenged the District Court’s determination that the defendants could not reasonably obtain evidence comparable in value to the destroyed evidence. Cooper, 983 F.2d at 931. The government argued the defendants could rely on experts familiar with the properties of laboratory equipment or, alternatively, could accept a jury instruction giving the defendants the benefit of any doubt. Cooper, 983 F.2d at 932.
The Ninth Circuit disagreed, finding, "[g]eneral testimony about the possible nature of the destroyed equipment would be an inadequate substitute for testimony informed by its examination”. Cooper, 983 F.2d at 932. The court further contrasted the comparable evidence available to the defendants there with the alternate evidence available in Trombetta. Cooper, 983 F.2d at 932. Because, in Cooper, the experts were prevented from examining the actual equipment as they were permitted to do in Trombetta, the Cooper court found the *501experts’ testimony did not permit the defendants to present an adequate defense. Cooper, 983 F.2d at 932.
Here, experts for the defense are faced with the same dilemma. They are asked to testify about machinery they are neither permitted to inspect nor even to become certified to operate. The Washington State Patrol allows only law enforcement personnel to operate or inspect the DataMasters and it denies requests by defense experts to access the machines. When defense witness Ms. Murren applied for a permit to operate the DataMaster, her application was denied based upon a "conflict of interest”. Murren testified the inability to access the machines limited her expertise and her ability to make informed decisions regarding the accuracy of a particular machine. Given these limitations, the experts’ testimony could be no more than mere speculation, and as in Cooper, it would be an inadequate substitute for testimony informed by its examination. Cooper, 983 F.2d at 932.
The insufficiency of this evidence is further supported by Trombetta itself. In Trombetta, the Court found the State’s failure to preserve the breath samples did not violate the defendants’ due process rights, in part, because the defense had access to the calibration records, as well as access to the Intoxilyzer machine itself. Trombetta, 467 U.S. at 490. The defendants thus had a means of impeaching the breath test results, other than retesting the original breath samples. Here, defense experts in Washington are prevented from accessing the DataMaster machines or even gaining certification on the machines. Thus, unlike the defendants in Trombetta, the Defendants in these cases do not have reasonably available means to obtain comparable evidence, and are thus precluded from mounting an adequate defense.
Our recent decision in State v. Straka, 116 Wn.2d 859, 810 P.2d 888 (1991) likewise supports this result. In Straka, this court held the State’s failure to generate and preserve invalid sample code messages when a breath testing machine was unable to complete a breath test did not violate a defendant’s due process rights in part because, "[a] defend*502ant still has the opportunity to attack the test results. Defendant may introduce evidence refuting the accuracy and reliability of the test reading”. Straka, 116 Wn.2d at 875. That was true, in part, because repair and maintenance records at issue here were still available to the defense in Straka. Under the majority’s decision today, defendants are effectively precluded from refuting the breath test results based on maintenance and repair records.
Not only does Trombetta require that defendants have access to alternate evidence, the opinion also makes it clear that defendants are not expected to jump through numerous hoops, procedural or otherwise, in order to obtain it. Instead, it must be obtainable "by other reasonably available means”. Trombetta, 467 U.S. at 489.
Here, the opposite is true. The Washington State Patrol has gone out of its way to ensure DWI defendants have no access to useful information. As discussed above, the State Patrol does not permit anyone other than law enforcement personnel to access the DataMaster machines. Nor is anyone other than law enforcement personnel permitted to become certified to use a DataMaster. Thus, notwithstanding the majority’s assertion that defendants can present expert testimony regarding the DataMaster machines, majority at 476, there are no real DataMaster experts accessible to the defense, other than those employed by the State Patrol.
In addition, the defense is sent on an apparent wild goose chase when it seeks the few records that are, by law, maintained. WAC 448-13-200 states that certain records are available on request. The only address provided in the WAC’s is that of the State Toxicologist’s office. WAC 448-13-210.2 Yet, it is actually the State Patrol that maintains the DataMaster records. The State Patrol, however, seems unwilling to part with the records. In one of the cases below, State ex rel. Dawson v. South Dist. Court, Snohomish Cy. cause 92-2-07095-1 (July 9, 1993), Petitioner Church wrote a letter to the State Patrol, attempting to obtain records for a *503DataMaster machine used in Snohomish County. In response to his letter, Church received form WSP-AS-327, denying Church’s request because "[t]he information [Church] requested is available ... at the South County Law Library”. Opening Br. of Pet’r (State ex rel. Dawson v. South Dist. Court) Ex. B.
There were, in fact, no records available at the law library. During the suppression hearing, the State Patrol first maintained there were no repairs records for the machine in 1992, even though Petitioner Church was directed to the law library to find the requested records. However, on the fourth day of the 7-day suppression hearing, when the records had potential evidentiary value for the State, the State Patrol produced the missing records, none of which had been stored in the Snohomish County Law Library. Surely, this conduct cannot comport with due process.
Ill
The opinions decided today set a dangerous precedent for a system of automated justice that is not only unfair, it jeopardizes a fundamental constitutional right to a fair trial. The records below establish a critical need for extrinsic evidence of DataMaster machine performance. The truth-finding function of a proceeding must allow for an inquiry into whether a machine was operating correctly when it issued a breath test result. I would find the failure to maintain repair and maintenance records violates due process under the state and federal constitutions, and would affirm the trial courts’ rulings suppressing the DataMaster test results.
Utter and Smith, JJ., concur with Johnson, J.
State v. Boland, 115 Wn.2d 571, 800 P.2d 1112 (1990) illustrates the breadth of analysis under factor four. In Boland, the defendant argued the warrantless search of his garbage, located in trash cans placed on the curb for collection, violated the fourth amendment to the United States Constitution and Const. art. 1, § 7. Becaúse the United States Supreme Court had recently held that such a search did not violate the federal constitution, the Boland court limited its analysis to the Washington State Constitution, using the six Gunwall factors. Boland, 115 Wn.2d at 575. See California v. Greenwood, 486 U.S. 35, 100 L. Ed. 2d 30, 108 S. Ct. 1625 (1988).
Applying factor four, preexisting state law, the court examined two local ordinances controlling trash disposal. The first, a Port Townsend ordinance, prescribed where to place trash cans for municipal pickup. Boland, 115 Wn.2d at 576. The second, a Seattle ordinance, made it "unlawful for anyone other than the owner of the trash can, or one authorized by the owner to place objects in the can, to remove its contents 'except for collection.’ ” Boland, 115 Wn.2d at 576 (quoting in part Seattle Municipal Code 21.36.100). Relying on these two ordinances, neither of which addressed a state constitutional basis, the court inferred an expectation of privacy in one’s own garbage and found the required support under factor four to review the case on independent state grounds.
WAC 448-13-210 provides the address for the State Toxicologist regarding inquiries for protocol, procedures, and authorized personnel.