Defendant appeals his sentence for hindering prosecution. The sole issue on appeal is whether the trial court erred in imposing a 36-month durational departure sentence. We conclude that the trial court erroneously interpreted and applied OAR 213-008-0002(l)(b)(J), the particular sentencing guideline on which the court relied in imposing defendant’s sentence. Consequently, we vacate the sentence for hindering prosecution and remand for resentencing.
Defendant pleaded guilty in connection with hindering the prosecution of Jesse Fanus, an individual who was suspected of murder. Defendant is a long-time friend of Fanus’s father and has known Fanus since he was an infant. Fanus was a fugitive from the police and a suspect in the murder of retired Marine Corps General Marion Carl and in the shooting of Carl’s wife in the course of a home invasion robbery. On June 29, 1998, defendant was driving through Roseburg and by chance saw Fanus on the street at a time that Fanus was in the vicinity of police officers, but apparently was not noticed by them. Defendant stopped his car, told Fanus he should get in, and drove him to Medford. Defendant did not know if Fanus had committed the crime. He suspected Fanus might have done so, however, because he knew that Fanus and Fanus’s brother had acquired a shotgun and he was concerned that they would get themselves into trouble with it. En route to Medford, Fanus told defendant that he had shot Carl with a shotgun and that he had disposed of the weapon. Defendant suggested that Fanus go to Los Angeles because it is a big city where “a person could get lost.” Once in Medford, defendant provided Fanus with some clothing and put him on a bus to California. One week later, California police arrested Fanus in suburban Los Angeles. While Fanus was on the run, defendant did not disclose Fanus’s whereabouts to anyone. When the police questioned defendant after apprehending Fanus, defendant initially denied any knowledge of the matter. He admitted his involvement only after learning that Fanus had told police that defendant had driven him to Medford.
*229Defendant was charged with and convicted of hindering prosecution. ORS 162.325.1 The sentencing court imposed an upward durational departure of 36 months’ imprisonment from the presumptive sentence of 15 to 18 months. In departing, the trial court relied on OAR 213-008-0002(l)(b)(J) (Factor J), which authorizes an upward departure if “[t]he degree of harm or loss attributed to the current crime of conviction was significantly greater than typical for such an offense.”
In arguing for a durational departure under Factor J, the state pointed to four facts or circumstances justifying the departure. First, the state urged that the crime committed by Fanus — aggravated murder — was a more serious felony offense than others that would support a hindering prosecution charge. Next, the state emphasized that defendant did not just provide Fanus with a ride away from the crime scene, but he took Fanus out of the county and “out of the net” where he might have been caught earlier. Third, the state contended that Fanus posed a graver risk to public safety than typically might be the case because he was an aggravated murder suspect and, according to the information available to police, he possessed a firearm. Finally, the state observed that, due to defendant’s offense, the search was expanded from a local to a state-wide search and finally to a national search that involved California and federal law enforcement authorities.
The trial court relied only on the third concern cited by the state — that is, the risk to the public created by the possibility that Fanus, while at large, would commit another murder:
*230“[Y]ou were giving [Fanus] a second breath to go down and potentially commit a similar crime or the same type of crime, and so you did put the public at great risk and you simply didn’t need to do it.
“So I think that the harm or loss caused by your actions were significantly greater than typical and that there will be a durational departure to 36 months * *
Defendant challenges the trial court’s departure under Factor J, arguing that Factor J authorizes a departure only if defendant’s crime actually caused harm significantly greater than typical for that crime, not if it only created a risk of such harm. The state responds that a risk of harm to the public, in the form of the potential for Fanus to commit more crimes, satisfies the language of the relevant departure factor. So framed, the issue turns on the meaning of the relevant sentencing guideline (Factor J), which presents us with a legal question that we decide without deference to its resolution below. See State v. Lark, 316 Or 317, 322, 851 P2d 1114 (1993).
The starting point for interpreting the guideline is, of course, its plain language. Factor J authorizes a departure if “[t]he degree of harm or loss attributed to the current crime of conviction was significantly greater than typical for such an offense.” (Emphasis added.) The text is phrased in the past tense. The guideline refers to a degree of harm or loss “attributed” to the current crime of conviction. It further refers to whether the harm or loss “was” significantly greater than typical for the crime. In that respect, the terms have a distinctly retrospective and completed-act focus. The same is true of the provision’s surrounding context. The other aggravating factors contained in the same rule are all directed to actual and completed harms that are a consequence of the crime committed, not to risk of harm in the form of prospective and theoretical future crimes.2 It may be that the legislature, had it considered the possibility, would have chosen to *231encompass risk of fixture harms in the equation.3 But, unlike other statutes and guidelines where the legislature expressly has referred to future or potential dangerousness, it plainly did not do so here. Compare ORS 163.150(l)(b)(B) (death sentence for aggravated mxxrder authorized where “there is a probability that the defendant would commit criminal acts of violence that would constitute a continuing threat to society”); OAR 213-008-0002(l)(b)(B) (departure sentence for person crime involving vulnerable victim authorized if the crime involved increased harm or “threat of harm”); OAR 213-008-0002(l)(b)(C) (same if person crime involving vulnerable victim involved “threat of or actual violence”).4
Moreover, Factor J refers to harm that is greater in “degree” than is typical for a particular crime. That language suggests quantitatively greater harm, not harm different in kind.5 A risk to public safety — which is the harm that the sentencing court identified — is not the harm that the crime of hindering prosecution seeks to prevent.
Hindering prosecution is a descendant of the common-law crime of accessory after the fact, which was an offense based on accessorial liability. At common law, acces-sorial liability “rested on the notion that one who helps an offender avoid justice becomes in some sense an accomplice *232in the original crime.” Model Penal Code and Commentaries § 242.3, 224 (Official Draft and Revised Comments 1985). One guilty of accessory after the fact was, in effect, derivatively liable for the underlying crime. Consistent with the notion of derivative liability, the accessory had to have knowledge that the principal committed the crime, the principal had to be tried first or jointly with the accessory, the principal’s conviction was a prerequisite to punishment of the accessory, and the accessory was subject to the same sentence as the principal. Id.
The Model Penal Code (MPC) broke “decisively” from the common-law view of the offense by rejecting its tradition of accessorial liability and adopting instead “the alternative theory of prosecution for obstruction of justice.” Id. at 224-25. The MPC therefore reformulated the crime as that of hindering apprehension or prosecution. It aimed the prohibition at the “purposeful efforts to aid another to evade justice” and did so “without regard to whether the person assisted in fact committed a crime and with penalties not invariably tied to those prescribed for the underlying offense.” Id. at 225.
In revising the Oregon Criminal Code in 1971, the legislature substantially adopted the MPC’s approach.6 The legislature repealed the former accessory statute (former ORS 161.230) and replaced it with hindering prosecution, codified as ORS 162.325. Although borrowing in part from the MPC, Oregon did not abandon accessorial liability altogether. Rather, it retained the requirement that the person aided be one “who has committed a crime punishable as a felony.” See ORS 162.325(1). Significantly, however, Oregon shifted the emphasis to the public’s interest in preventing the obstruction of justice. The legislature changed the mental state required for hindering prosecution by deleting the requirement that a defendant know that the person he or she aided in fact had committed a felony. Instead, the mental *233state required is the intent to hinder apprehension or prosecution, thus aiding the offender in “escaping justice.” See generally Commentary to Oregon Criminal Code of 1971, § 162.325,106 (1975).
As that history reflects, hindering prosecution in Oregon is still based in part on accessorial liability for the crime committed by the aided felon, but the penal goal of the offense is to prevent the obstruction of justice. Consistent with that objective, hindering prosecution does not require that an offender know that he or she is aiding someone who in fact committed a crime; it does not require that the person aided have recidivist tendencies; it does not require actual success in avoiding prosecution.7 To be sure, if the hindering effort is successful, then a felon may go free and may commit crimes in the future. That possibility, however, is at most a potential collateral consequence of the crime’s commission. It is not a harm that inheres in the crime’s commission or that forms the rationale for criminalizing the conduct.
The trial court’s departure under Factor J, therefore, reflects two legal errors. First, the court incorrectly considered harm that was theoretical only, rather than a harm or loss that actually occurred because of defendant’s crime. Second, the court focused on the wrong harm altogether — that is, the court did not assess whether defendant’s crime interfered with public justice in a way that was significantly greater than is typical for a hindering offense. The departure sentence therefore is not sustainable based on the rationale offered by the trial court for imposing it.
We cannot say, however, that there is no basis on which a departure sentence might be authorized in this case. The appropriate disposition, therefore, is to vacate the sentence for hindering prosecution and to remand for resentencing. See ORS 138.222(5); see generally State v. Edson, 329 Or 127,139, 985 P2d 1253 (1999) (in restitution context, statute *234requires remand for resentencing when there remain bases on which trial court might be able to exercise discretion to impose restitution).
Conviction affirmed; sentence vacated; remanded for resentencing.
ORS 162.325 provides, in part:
“(1) A person commits the crime of hindering prosecution if, with intent to hinder the apprehension, prosecution, conviction or punishment of a person who has committed a crime punishable as a felony, * * * the person:
“(c) Provides or aids in providing such person with money, transportation, weapon, disguise or other means of avoiding discovery or apprehension; * * *
‘(2) Hindering prosecution is a Class C felony.”
Other aggravating factors in the same administrative rule include that the crime involved deliberate cruelty to the victim; knowledge of a victim’s particular vulnerability; use of a weapon; a violation of public trust or responsibility; multiple victims; permanent injury to the victim; and a discriminatory animus in targeting a particular victim. See generally OAR 213-008-0002(l)(b).
The guidelines were adopted first by the Oregon Criminal Justice Commission and then approved by the legislature. See generally ORS 137.667 and ORS 137.669.
Assessing a person’s future criminal propensities might be particularly problematic if, as the trial court assumed, the focus is on the person aided, rather than the criminal propensities of the person being sentenced. The trial court would be enhancing this defendant’s sentence based on a crime not adjudicated by the court and based on the criminal proclivities of an individual not before the court. Thus, in this case, the only basis on which the trial court made an assessment of Fanus’s future dangerousness was through an assumption that, because Fanus had committed a crime in the past, he will commit a like crime in the future, which is an assumption that for many purposes the law does not permit. See generally OEC 404 (other crimes evidence not admissible to show a person’s tendency to commit crimes). We are unwilling to interpret the guidelines to authorize a sentencing court to enhance a defendant’s sentence based on predictions of someone else’s future dangerousness without an unequivocal expression of the legislature’s intent to do so.
The most closely applicable ordinary meaning of “degree” is “a grade or point observed in a measuring or estimating of an action, relation, [or] state of being * * í= ” Webster’s Third New Int’l Dictionary, 594 (unabridged ed 1993).
Oregon also borrowed the modifications to the MPC provision that had been adopted by New York and Michigan.
Actual obstruction of a prosecution is not an element of the crime. Under the former crime of accessory after the fact, the state’s burden of proof could be satisfied if it showed that a defendant engaged in a prohibited act with the intent to interfere in apprehension or prosecution. Success was not an element. State v. Clifford, 8 Or App 494, 491 P2d 1195,495 P2d 49 (1971), rev’d 263 Or 436,502 P2d 1371 (1972) (finding insufficient evidence to support guilty verdict). The statute has not changed in that regard.