HASELTON, C. J.
Defendant was charged with multiple counts of sexual abuse. ORS 163.427. After a first trial resulted in a mistrial, the state reprosecuted and defendant moved to dismiss based on (1) preindictment delay, invoking the Fifth and Fourteenth Amendments to the United States Constitution, and (2) lack of speedy trial, invoking former ORS 135.747 (2011), repealed by Or Laws 2013, ch 431, § 1, Article I, section 10, of the Oregon Constitution, and the Sixth and Fourteenth Amendments to the United States Constitution. In a separate motion, defendant also sought dismissal on double jeopardy grounds, ORS 131.515, Article I, section 12, of the Oregon Constitution, and the Fifth and Fourteenth Amendments. As detailed below, the trial court granted the former motion, denied the latter motion, and entered a judgment of dismissal.1
The state appeals, arguing that the trial court improperly conflated pre- and post-indictment delay in its analysis supporting the dismissal, and that, in any event, with respect to the preindictment delay, defendant failed to establish that the state was culpable for the delay and that he was actually prejudiced by the delay.
Defendant responds, arguing, inter alia, that the state was negligent or indifferent to the delay and that he was actually prejudiced because, as a result of the delay, an important impeachment witness was unavailable. Defendant also cross-assigns error to the denial of his motion to dismiss based on double jeopardy.2
[332]*332As amplified below, we affirm based on our conclusion that the preindictment delay violated defendant’s right to due process,3 because the lengthy delay was unjustified and defendant was actually prejudiced by the loss of the opportunity to impeach the victim through the testimony of the principal investigating detective who was unavailable because of that delay. Accordingly, we do not address the state’s arguments challenging the trial court’s conclusion that defendant’s constitutional and statutory speedy trial rights were violated. Similarly, given our analysis and disposition, we need not reach defendant’s cross-assignment of error pertaining to the trial court’s denial of his motion to dismiss on double jeopardy grounds.
We review a trial court’s ruling on a motion to dismiss for errors of law. State v. Davis, 345 Or 551, 564-65, 201 P3d 185 (2008), cert den, 558 US 873 (2009). We are bound by the trial court’s findings of historical fact that are supported by evidence in the record. State v. Ehly, 317 Or 66, 74-75, 854 P2d 421 (1993). To the extent that the trial court did not make express findings, we resolve disputed facts consistently with the trial court’s ultimate conclusion. Ball v. Gladden, 250 Or 485, 487, 443 P2d 621 (1968).
I. BACKGROUND
The background facts of this case are undisputed and derived from testimony in a trial that took place in February 2011.4 Defendant is the step-grandfather of the alleged victim, S, who, between the ages of eight and 12, lived with defendant and her maternal grandmother (defendant’s wife) at their home in Sheridan. In August 2004, when S was 12 and still living with defendant and her grandmother, S disclosed to her mother that she believed that defendant had sexually abused her. On August 24, 2004, S reported her allegations to Yamhill County Sheriff [333]*333Corporal Ludwig, who took an initial report and then assigned the case to Detective Rosario. That same day, S moved out of defendant’s home. On August 25, Detectives Rosario and Carelle interviewed S and her mother at the sheriffs office.5 On September 2, S was examined and interviewed at a child abuse assessment center. A center staff member, in turn, reported S’s disclosures to Rosario, who then authored an investigative report. As explained below, that report by Rosario is central to defendant’s contention that he suffered actual prejudice as a result of Rosario’s subsequent unavailability. Although Carelle was present at the initial interview, he was not in charge of the investigation, and he did not author any reports regarding the investigation. After writing his report, Rosario did not engage in any further investigative activity. Consequently, the investigation halted in early September 2004, mere days after it commenced.
On March 28, 2006,19 months after S reported the alleged abuse, a Yamhill County Multidisciplinary Child Abuse Team (MCAT)6 met to discuss her allegations. After that meeting, the Yamhill County District Attorney’s Office sent a prosecution memo to Rosario with a “prosecution declined” box checked and the following explanatory note:
“No response to requested follow-up. Victim is unable to be located. If further follow-up is produced, please feel free to re-refer case. If you can locate victim, please advise.”
The same memo notes that the case was discussed at an MCAT meeting on April 4, 2006, but no further action was suggested or taken. In March and April 2006, S was living with her mother in Dallas, Oregon (in neighboring Polk County) and attending public middle school.
In 2007, S moved back into defendant’s home in Sheridan and attended public high school. On November 21, 2008, an anonymous caller contacted the Department of [334]*334Human Services (DHS), and reported that “[S] had said [defendant] tried to molest her.” Three days later, Yamhill County Sheriffs Office Detective Geist (who was unaware of the previous allegations and investigation involving defendant) and a DHS service worker visited defendant’s house to check on S, who was then 16 years old. S did not report any abuse at that meeting.
The next day, November 25, Geist discovered a record of the case, including Rosario’s report. Geist “asked around” about Rosario — who had left his position at the Yamhill County Sheriffs Office in March 2008 — and learned that Rosario was reportedly living in Puerto Rico. Geist did not attempt to locate or contact him there.7 Geist testified:
“I found in that report that Detective Rosario had spoken with [S] and her mother along with Sergeant Carelle from our office and then [S] had been seen at * * * the child abuse assessment center here in town. I believe [S] was seen in early September of 2004. Detective Rosario’s report indicated that there would need to be follow-up done to complete the case. I did not find any other follow-up in that case in [our case records management system], so I had records pull the case file and found that there was no follow-up done in the case file, either.
«* * * * ‡
“Being discussed at MCAT is not follow-up on an investigation. *** I discuss all my cases at MCAT; that doesn’t mean I followed up on it just because I discussed them with MCAT. That is not a follow-up. Follow-up is when I actually go out and do something on the case.”
After learning that the 2004 investigation remained unresolved, Geist decided to contact S at school. At that interview, S not only reaffirmed her allegations from 2004, but also reported other instances of abuse and inappropriate behavior by defendant when she had lived with him between [335]*335the ages of 10 and 12 years old. Some of those incidents were not included in Rosario’s report. S stated that no new incidents of abuse had occurred after she had moved out of defendant’s home in August 2004. Geist contacted defendant, who categorically denied S’s allegations.
On June 17, 2009, nearly five years after S first reported the alleged abuse, the state arrested defendant and indicted him on multiple counts of sex abuse. Defendant entered a not-guilty plea, and the case proceeded to trial in February 2011. At that trial, S testified to incidents that were not included in Rosario’s 2004 report. For example, S testified about an incident where defendant laid on top of her and kissed her while she was laying on her grandparents’ bed. When defense counsel asked her whether she told Rosario about that incident, she responded, “I don’t remember.” In another instance, S testified that defendant kissed her on the stairs in his garage and tried to put his tongue in her mouth. Defense counsel asked her whether she told Rosario about that incident, and she responded, “I don’t remember.”
That first trial ended in a mistrial.8 The state elected to reprosecute defendant and, in June 2011, before a second trial, defendant moved to dismiss for lack of speedy trial and delay constituting denial of due process. In particular, defendant contended that, in light of the lengthy pre-indictment delay, “[a]ny further prosecution * * * would deny him Due Process of law.” That was so, defendant argued, because “no substantive investigation took place” after the initial interview and assessment in August and September 2004, and the state provided no justification for delaying the indictment. Defendant also asserted that further prosecution would deny him his statutory and constitutional right to a speedy trial. In advancing that contention, defendant combined the 58-month period between S’s report and his indictment with the 26-month period after the indictment and before the second set trial date. Defendant conceded [336]*336that some of the post-indictment delay was attributable to his motions to continue, but he argued that the remainder of the delay was attributable to the state.
The trial court granted that motion to dismiss after finding that the total preindictment delay was 58 months9 and “entirely attributable to the state.” In that regard, the court found that “[t]here was a significant period in the pre-indictment stage where no investigation was taking place.” In response to the state’s proffered reason for the delay— that S was unable to be located — the court found that she “could have been located through family, associates, or school with little effort.” Accordingly, the court determined that the delay was unjustified and “a result of negligence or indifference by law enforcement in following through on an investigation that might easily have located the victim and allowed further pursuit of the case.”
With respect to actual prejudice to defendant as a result of that delay, the court determined:
“As a result of the delay, defendant and other witnesses would likely lose memory of events. During the delay, both pre-indictment and post-indictment, the victim herself lost memory as to events as evidenced by her answering at the February trial, T don’t remember,’ about 30 times.
“Mr. Rosario, the primary investigator, is not available for his testimony or to recount statements of [S] on prior occasions. Cross-examination of Mr. Rosario could also be especially important in this type of case to test the conformance of the investigation to the Oregon Interviewing Guidelines, intended, among other things, to assure the integrity and non-leading nature of questions of the victim.”
After making those findings and observations, the trial court concluded:
“In this case, harm to the defense can be identified in some specific ways: loss of memory of events by the defendant, victim, and others, and the unavailability of Mr. Rosario. There are undoubtedly other hard-to-identify ways in which the delay has compromised the ‘reliability of atrial.’”
[337]*337(Emphasis in original.) Considering the pre- and post-indictment delay together, the trial court concluded that the “overall delay” violated defendant’s right to due process and his right to a speedy trial. The court subsequently entered a judgment of dismissal, from which the state appeals.
On appeal, the state argues that the trial court improperly combined pre- and post-indictment delay in ruling on defendant’s motion to dismiss. With respect to the preindictment delay alone, the state contends on appeal that defendant failed to establish that the state is culpable for the delay and that he was actually prejudiced by the delay.
Defendant remonstrates that the state negligently delayed bringing the indictment, as evidenced by the fact that no investigative activity took place for over four years after S reported her allegations of abuse to the police in 2004. Defendant contends that he was actually prejudiced by the unavailability of Rosario, an important impeachment witness, whose unavailability was caused by the preindictment delay.
II. ANALYSIS
At the outset, we agree with the state that the trial court improperly combined pre- and post-indictment delay in ruling on defendant’s motion to dismiss. Preindictment delay implicates a defendant’s right to due process, and is analyzed separately from post-indictment delay, which implicates a defendant’s right to a speedy trial. See State v. Endres, 196 Or App 197, 200-03, 100 P3d 784 (2004) (explaining that distinction). Nevertheless, as the state implicitly recognizes, our conclusion that the trial court erred in considering the pre- and post-indictment delay together does not materially affect our review of whether the trial court properly concluded that defendant’s right to due process was violated by preindictment delay alone. That is so because the trial court rendered findings as to the causation of the preindictment delay and the prejudice to defendant as a result of that delay; and our task is to determine whether, under the applicable due-process construct, those determinations in combination compelled dismissal.
State v. Stokes, 350 Or 44, 248 P3d 953, cert den, ___US___, 132 S Ct 343 (2011), frames and guides our [338]*338consideration. In Stokes, the court prescribed the controlling test:
“To demonstrate that preindictment delay violated the federal Due Process Clause, a defendant must show that the delay actually prejudiced the defendant and that the government culpably caused the delay. A court must weigh the government’s reason for the delay against the prejudice to determine whether the delay violated our society’s fundamental conceptions of justice, fair play, and decency.”
350 Or at 64; see also Davis, 345 Or at 574 (a defendant must establish some causal link between the delay and the asserted prejudice).10
We begin our analysis of preindictment delay by noting that there is no dispute that — as the state acknowledged at oral argument on appeal — Rosario was unavailable as a result of the preindictment delay in this case. We accept and appreciate the state’s concession in that regard, which is well founded, given that Rosario left the Yamhill County Sheriffs Office in March 2008 — more than three years after the initial report of abuse and 14 months before defendant was indicted.
With that predicate causal nexus established, consistently with the conjunctive test set out in Stokes, we proceed to examine in turn whether the state was culpable for the delay and, if so, whether defendant was actually prejudiced by the delay. Finally, we balance the asserted prejudice against the state’s reasons for the delay.
As we will explain, we conclude that the state was unjustified and negligent in delaying the indictment, and [339]*339that defendant’s defense was actually prejudiced as a result. We also conclude that, on balance, the prejudice to defendant outweighs the state’s reason for the delay. Accordingly, the trial court did not err in concluding that the preindictment delay violated defendant’s right to due process and, consequently, dismissing the charges against him.
A. Culpable Delay
We begin with the state’s proffered reason for the delay. Stokes, 350 Or at 64. The importance of the state’s activity during the delay, and the attendant reasonableness of its justification for the delay, adds weight to the state’s side of the due-process balance, while weak or unreasonable explanations for the delay will reduce the weight on that side. Put another way, as the state’s culpability in the delay increases, the weight of its justification decreases. “Recklessness or even negligence on the government’s part” may be weighed in due-process balancing. Id. at 57.
The state does not assert on appeal — as it did before the trial court — that the 5 8-month preindictment delay was justified, specifically, by difficulties in locating S. Instead, citing United States v. Lovasco, 431 US 783, 796, 97 S Ct 2044, 52 L Ed 2d 752 (1977), the state contends that the delay was, generally, excusable as “investigative delay” and, as such, did not deprive defendant of due process. That contention is unavailing in that the state’s characterization of the preindictment delay here as “investigative delay” indiscriminately equates (a) delay that is attributable to case-specific difficulties in completing an investigation and initiating prosecution and (b) delay that is either unexplained or is the product of inattention or indifference in undertaking and pursuing investigation or in initiating prosecution. Under Lovasco, only the former is constitutionally justified delay. Conversely, here, the preindictment delay was almost exclusively of the latter type.
Lovasco itself, as well as Davis and Stokes, illustrate that distinction. In Lovasco, federal agents investigated the defendant for possessing and selling firearms that had been stolen from the US mails. 431 US at 784. The defendant admitted to possessing and selling five of eight stolen guns [340]*340in the first month of the investigation. Id. at 785. However, an indictment was not filed against the defendant until 18 months later. The defendant moved to dismiss, arguing that he was prejudiced because two potential witnesses had died during the preindictment delay (one had died eight months after the alleged crimes occurred and the other died more than a year after). The district court dismissed the indictment and the circuit court affirmed.
On review, the United States Supreme Court reasoned that “proof of prejudice is generally a necessary but not sufficient element of a due process claim,” and that “the due process inquiry must consider the reasons for the delay as well as the prejudice to the accused.” Id. at 790. The Court explained that the Fifth Amendment does not operate to force a prosecutor to indict a suspect upon obtaining clear evidence of guilt: “Rather than deviating from elementary standards of ‘fair play and decency,’ a prosecutor abides by them if he refuses to seek indictments until he is completely satisfied that he should prosecute and will be able promptly to establish guilt beyond a reasonable doubt.” Id. at 795. Accordingly, the Court held that “to prosecute a defendant following investigative delay does not deprive him of due process, even if his defense might have been somewhat prejudiced by the lapse of time.” Id. at 796.
In so holding, the Court observed that, in that case, “the investigation continued during the time that the Government deferred taking action against [the defendant] .” Id. Specifically, the Court accepted the government’s representations that the delay was caused by the government’s efforts to identify other possible participants in the crime under investigation. Id. Under those circumstances, in which the criminal “investigation continued,” the Court concluded that there was no culpable government conduct- — i.e., the government was justified in the delay — and, therefore, the defendant’s right to due process was not violated. Id.
Davis — in which the court concluded that an 11-year period of preindictment delay did not compel dismissal— also exemplifies justifiable investigative delay. We necessarily recount its circumstances in some detail. In Davis, the defendant’s ex-girlfriend and a male companion were shot [341]*341and killed at a Portland motel on the evening of November 2, 1991. The victims died, from wounds from a .45-caliber weapon, sometime between 8:20 p.m. and 10:20 p.m. 345 Or at 553.
A police detective interviewed the defendant shortly after the murders, id. at 555, and the defendant admitted that he had seen the female victim on the day of the murder, but proffered an alibi for that evening: According to the defendant, he had spent the evening at home, and his friend and housemate, Bynum, had called him between 8:00 p.m. and 8:30 p.m. to discuss that night’s professional basketball game during halftime. Id. at 556. Then, in the defendant’s account, he had left his apartment between 10:00 p.m. and 10:30 p.m. to buy beer and cigarettes at a Safeway store and, when he returned to the apartment, Bynum was there. Id. Finally, according to the defendant, he had spent the later part of the evening with other friends, including Foreman, Lowery, and Payne, and, with those men he had visited a nightclub and two restaurants, returning to his apartment between 3:00 a.m. and 4:00 a.m. on November 3. Id. at 557.
Portland police subsequently interviewed Bynum, Foreman, and Lowery, who all described the events of the evening consistently with the defendant’s alibi. Id. at 557-58. Police also attempted, but failed, to contact Payne. Id. at 558.
With the defendant’s alibi corroborated, and no other inculpatory evidence against him, the investigation “stalled” until April 2002,11 when an inmate at the Oregon State Penitentiary, where the defendant was incarcerated on unrelated convictions, informed detectives that the defendant had admitted to the murders. Id. at 560. Soon thereafter, detectives again interviewed Bynum, who told detectives that the defendant had admitted the killings to him after the initial investigation. Id. at 561.
Those developments prompted Portland police to locate and contact Payne, who relayed especially damning [342]*342information about the defendant’s actions during the night of the murders, including that, while the defendant had been loading a .45-caliber pistol, he had “ranted” that he was going to kill the female victim, and that between 9:00 p.m. and 10:00 p.m. the group of men had left the defendant at a motel where he had expected to find the victims. Id. at 562. Payne also told the detective that the defendant had threatened to kill him if he talked to the police. Id. at 563.
Detectives then contacted Lowery, who, during the 1991 investigation, had corroborated the defendant’s alibi. Id. Lowery confirmed that the defendant had left the group of men near the motel where the murders occurred. Id. Lowery also told detectives that he and Foreman had agreed to tell police the truth about other events on the evening of the murders, but to omit the trip to the motel, and the defendant had later thanked them for supporting his alibi. Id. at 563-64. Lowery and Payne both testified before a grand jury in November 2002, which indicted the defendant on eight counts of aggravated murder. Id. at 564.
Before trial, the defendant moved to dismiss, contending that the 11-year preindictment delay had violated his right to due process. Specifically, the defendant posited that evidence had been lost, destroyed, or not timely located due to the delay, including testimony from other guests at the motel on the night of the murders, crime-scene evidence, defendant’s clothes from the night of the murders, and interviews of Safeway personnel where defendant had claimed that he had purchased beer and cigarettes. Id. 565, 573-74. The trial court denied the defendant’s motion, and a jury found the defendant guilty on all counts; he was sentenced to death. Id.
On direct review, the Oregon Supreme Court, applying the federal due process analysis, examined the nature and causation of the preindictment delay. Integral to that assessment was the trial court’s finding that “the investigation stalled in 1991, shortly after the murders, and then stalled again in 1996 — each time due to [the] defendant’s skill in constructing an alibi.” Id. at 577-78 (citing with approval the trial court’s description of the preindictment delay as the product of “a ‘conspiracy of circumstances’ caused by the [343]*343‘the solidity of,’” inter alia, the defendant’s alibi). The court concluded: “[T]he trial court’s findings establish that the lengthy preindictment delay in [the] defendant’s case was the result of the state’s good faith efforts to investigate the case and to gather needed evidence” to provide “sufficient probable cause” to charge the defendant with the murders. Id. at 578. Accordingly, “the delay was for a constitutionally permissible reason.” Id.
Finally, in Stokes, a grand jury declined to indict the defendant on allegations of sodomy and sexual abuse of his adult male friend. 350 Or at 46. During an initial investigation, two other men claimed to have been sexually abused by the defendant; but those allegations did not result in charges at that time. Id. That same year, the defendant was incarcerated in California for unrelated convictions. The prosecutor in Oregon, believing that the defendant would remain incarcerated in California for the remainder of his life, informed the police department that the case was closed. Consequently, the police destroyed evidence related to the case, including the first victim’s 9-1-1 call, the contents of a rape kit, the victim’s clothing, and towels and clothing seized from the defendant’s house. Id. at 47.
Eight years later, a different Oregon prosecutor learned that the defendant’s California conviction had been reversed and that, consequently, he would be released from custody. Id. That prosecutor reopened the case and resubmitted the previous charges to the grand jury along with new charges of sodomy and sexual abuse of one of the other earlier-reported victims. The grand jury indicted the defendant on all of the charges. Id. at 47. After a jury trial, the defendant was convicted on all counts. Id.
On review, in addressing the defendant’s due process challenge, the Supreme Court determined that the pre-indictment delay was for constitutionally permissible reasons. In so holding, the court emphasized that, “[w]hile the case was open, the state was investigating and moving the case forward,” and that the state initially closed the case based only “on the good-faith belief that [the] defendant would serve a life sentence” in California; it was only when circumstances changed, and that apparently reasonable [344]*344assumption proved to be incorrect, that the state promptly reopened the Oregon investigation. Id. at 62. While acknowledging that those circumstances could not “be classified as investigative delay” in Lovasco’s rubric, the court nevertheless concluded that the totality of circumstances did “not demonstrate * * * government culpability” of the sort necessary to establish a due process violation based on preindictment delay. Id. at 62-63.
The circumstances of the preindictment delay in this case are materially, decisively different from those in Lovasco, Davis, and Stokes. The delay here was neither excusable “investigative delay” nor otherwise “nonculpable” delay (as in Stokes).
Unlike in Lovasco, where the court expressly determined that the investigation continued during the delay while law enforcement searched for additional suspects, 431 US at 796, the trial court here found that “[t]here was a significant period in the preindictment stage where no investigation was taking place!’ (Emphasis added.) Specifically, investigative activity inexplicably ceased in September 2004, almost immediately after it began, and no further investigation occurred until DHS received an anonymous call in November 2008. Further, here, unlike in Lovasco, there were no other potential participants or witnesses for the state to locate, delaying the course of the investigation.
Davis is even more dramatically distinguishable. As noted, the delay in investigation there was the product of the defendant’s own calculated conduct in orchestrating a “conspiracy of circumstances” to fabricate an alibi ostensibly corroborated by multiple witnesses. Davis, 345 Or at 577. Here, there is no suggestion that defendant somehow impeded the prompt completion of investigation and initiation of prosecution; nor does the record disclose any other significant investigative difficulties, much less any that would justify very substantial preindictment delay. Thus, the delay here is not “the result of the state’s good faith efforts to investigate the case and to gather needed evidence.” Id. at 578.
Finally, Stokes is inapposite. As noted, there, the state actively pursued the investigation so long as the case [345]*345was open, and closed the case only after making a good-faith determination that, given the defendant’s anticipated incarceration on other charges for the rest of his life, continued investigation and prosecution “was not a good use of public resources.” 350 Or at 63. In contrast, here, the state has made no assertion that the delay was justified by calculated considerations of the best use of public resources: The state has provided no rational, good-faith justification for why the case against defendant languished for years. Instead, as the trial court found, that delay was “a result of negligence or indifference by law enforcement in following through on an investigation that might easily have located the victim and allowed further pursuit of the case.” Accordingly, defendant established that the state was culpable in delaying the indictment. See Stokes, 350 Or at 57 (explaining that “[Recklessness or even negligence on the government’s part” may be weighed in due-process balancing).
B. Actual Prejudice
We turn to the second component of the due process balancing construct — actual prejudice. We begin with two overarching observations that substantially inform our analysis: First, as noted above, 262 Or App at 338, it is uncontroverted on appeal that, but for the preindictment delay, Rosario, the principal investigating detective, would have been available to testify. Second, as described more fully immediately below, defendant has identified with particularity the asserted actual prejudice resulting from Rosario’s unavailability. Those circumstances, individually and collectively, distinguish this case from, e.g., Davis and Stokes, where the Oregon Supreme Court determined that the asserted prejudice, balanced against lengthy, but justified, preindictment delay, did not compel dismissal.
The gravamen of defendant’s asserted actual prejudice resulting from Rosario’s unavailability as a witness at trial is: (1) In this prosecution for sexual abuse, in which there were no witnesses to the alleged conduct and no corroborating physical evidence, the credibility of the complainant, S, is pivotal; (2) Rosario’s investigative report describes certain conduct that S recounted during her August 2004 interview with him and during her contemporaneous abuse [346]*346assessment; (3) during the February 2011 trial, which ended in a mistrial, S testified as to multiple instances of abuse, and types of abuse and inappropriate behavior by defendant that were not referenced in Rosario’s report;12 (4) that disparity between the content of Rosario’s contemporaneous report and recounting of S’s allegations and S’s trial testimony (and putative testimony in the putative second trial) is indicative of substantial inconsistency on S’s part; and (5) testimony from Rosario as to the diligence and thoroughness of his investigative inquiry and the comprehensive accuracy of the content of his report would materially buttress the inference of inconsistency, subverting S’s credibility.
With respect to the final proposition, defendant asserts particularly that Rosario was uniquely qualified to testify about his interviews with S — what he asked her, how she responded, and which of her statements he did or did not include in his report. As defendant’s counsel (who was also his trial counsel) explained during oral argument before this court:
“My questions to [Detective Rosario] would be:
“[Defense Counsel]: Detective Rosario, you have received training in writing reports, isn’t that true?
“[Rosario]: Yes.
“ [Defense Counsel]: And, in that training, you learned to write complete reports, isn’t that true?
“[Rosario]: Yes.
“[Defense Counsel]: You learned to write accurate reports, isn’t that true?
“[Rosario]: Yes.
“[Defense Counsel]: You learned to write full reports, isn’t that true?
“[Rosario]: Yes.
“[Defense Counsel]: And, when you receive an answer that requires a follow-up question, you ask that follow-up question, don’t you?
[347]*347“[Rosario]: Yes.
“[Defense Counsel]: And is that what you did in this easel
“[Rosario]: Yes.”
(Emphasis in original.) Defense counsel further explained the ultimate, practical implications of Rosario’s testimony, given S’s putative testimony:
“After [S] repeatedly said, ‘I don’t remember’ [whether she told Rosario about certain allegations], then, [the defense] brings Rosario back and goes through those [numerous] instances with him, where he says, ‘No, she didn’t tell me that. No, she didn’t tell me that.’ The jury hears that and, in their deliberations would wonder, ‘Well, why didn’t she tell him that
(Emphasis added.)
The state does not dispute that, at the time of the indictment and at trial, S alleged and recounted conduct that was not included in Rosario’s report. Instead, invoking Davis and Stokes, the state asserts that the alleged prejudice resulting from Rosario’s unavailability either does not exist or is “speculative” at best. We disagree: Once again, the circumstances of this case are decisively different from those in either Davis or Stokes.
In Davis, as noted, the defendant posited that evidence had been lost, destroyed, or not timely located due to the delay, including testimony from other guests at the motel on the night of the murders, crime-scene evidence, defendant’s clothes from the night of the murders, and interviews of Safeway personnel. 345 Or at 565, 573-74. In rejecting that claim of actual prejudice, the court first observed that, because much of the allegedly “unavailable” evidence “likely became unavailable weeks or months after the crimes were committed,” the defendant had failed to demonstrate that, but for the constitutionally impermissible delay, the evidence would have been otherwise available. Id. at 574 (emphasis added). The court further reasoned that the evidence that the defendant identified was equally as likely to harm as to aid his defense. Accordingly, the court concluded, the prejudice that the defendant asserted was “purely speculative.” Id. at 575-76.
[348]*348In Stokes, the defendant contended that his defense was actually prejudiced by the destruction of evidence and the death of three potential witnesses, but he did “not offer a specific reason” why the loss of that evidence was prejudicial to his defense. 350 Or at 54-55, 60. The court reasoned that, without knowing the quality of the destroyed or unavailable evidence, “[the] defendant [could] only speculate that it might have helped his defense.” Id. at 60.
Here, in contrast to Davis, it is undisputed that, but for the lengthy preindictment delay, Rosario would have been available as a witness at trial. See 262 Or App at 338. Further, and most significantly, unlike in Davis and Stokes, the nature and content of the “missing” or “lost” evidence, Rosario’s putative testimony, and its utility to the defense, is far from “speculative.” It is hardly “speculative” to assume that an experienced detective would confirm the thoroughness of his or her interview with the complainant in a sexual abuse investigation and the comprehensive accuracy of the consequent investigative report. Nor is it speculative to posit that such first-person testimony by the principal investigating detective — as opposed to mere admission of the report itself — would likely powerfully buttress an implication of material inconsistency by the complainant.13
The state remonstrates, nevertheless, that defendant could have achieved the same result by cross-examining the other detective, Carelle — and, thus, the defense was not actually prejudiced by Rosario’s unavailability. As defendant counters, “It is not the same.” Although Carelle was present at an interview, he did not communicate with the sex abuse assessment center staff, and he did not author the investigative report from which S’s later allegations were omitted. It is the discrepancy between Rosario’s report and S’s testimony that is crucial to defendant’s impeachment theory; Carelle’s testimony is not an equivalent substitute.
[349]*349We conclude our discussion of “actual prejudice” by reiterating the centrality of the assessment of credibility— and, derivatively, of impeachment evidence — in these circumstances. As we have emphasized in a host of post-Southard decisions, in cases of unwitnessed sexual abuse in which there is no corroborating physical evidence, the complainant’s credibility is often decisive.14 Further, as the Oregon Supreme Court has admonished in an analogous context, “where the impeached witness is the sole witness on a given issue and there is no corroborating evidence, the interests of a fair trial require that the adverse party be given ample opportunity” to impeach that witness. State v. Hubbard, 297 Or 789, 800-01, 688 P2d 1311 (1984) (emphasis added).15
Defendant was actually prejudiced by the lack of opportunity to impeach S by way of Rosario’s testimony.
C. Conclusion
Finally, we must weigh the established actual prejudice against the state’s reason for delay. Stokes, 350 Or at 57. In Davis, the court explained that the due process balancing test “entails a sliding scale in which the more substantial the actual prejudice to the defendant becomes, the greater the reasonableness and the necessity for the delay must be.” 345 Or at 576 (internal quotation marks omitted). [350]*350There, because the defendant’s claims of prejudice were “entirely speculative,” the court gave no weight on the side of prejudice. In contrast, the state’s justification for the delay— the defendant’s alibi derailed the investigation and police discovered new inculpatory evidence years after the initial investigation — weighed in the state’s favor. On balance, the state’s reason for delay outweighed the weightless, speculative prejudice. Id. at 578.
Similarly, in Stokes, the court concluded that the defendant “showed either no actual prejudice, or only the slightest actual prejudice,” and the state was justified in the delay. 350 Or at 63-64 (explaining that the state “closed the case based on a conclusion that [the] defendant’s California sentence minimized his threat and, in light of that, pursuing defendant’s prosecution in Oregon was not a good use of public resources”). The court concluded that the state’s reason for delay outweighed any slight prejudice. Id.
Here, defendant is actually prejudiced by the absence of the opportunity to impeach S with testimony from Rosario, who is unavailable as a result of the state’s delay in bringing the indictment. On the other end of the scale, the state’s delay in bringing the indictment was due solely to negligence or inexcusable inattention in pursuing the investigation. In that calculus, we give the state’s reason for delay — which is, effectively, no justification at all, specifically with respect to the period between September 2004, when the investigation inexplicably went dormant and November 2008 — no weight. Accordingly, on balance, the actual prejudice to defendant, which, in this credibility-predicated case is substantial, outweighs the state’s reason for the delay. We affirm the trial court’s dismissal on the ground that the preindictment delay in this case violated defendant’s right to due process. Given that disposition, we dismiss defendant’s erroneously denominated “cross-appeal,” see 262 Or App at 331-32 n 2, as moot.16
Affirmed on appeal; cross-appeal dismissed as moot.