This case is a powerful indictment of the criminal justice system. Our social and penal policies are failing to alleviate alcohol abuse on Indian reservations and the crime to which it gives rise. These problems cry out for treatment, not simply more prison time.
Vernon Lee Bad Marriage, Jr. (“Bad Marriage”) is a member of the Blackfeet Indian Tribe with an extensive history of alcohol abuse and a lengthy criminal record. He was convicted of assault resulting in serious bodily injury, in violation of 18 U.S.C. §§ 113(a)(6) and 1153(a), and sentenced under the U.S. Sentencing Guidelines. The District Court departed upward from the applicable sentencing range on the grounds that Bad Marriage’s criminal history score did not adequately reflect the seriousness of his past criminal history and the likelihood that he would commit other crimes. Because we hold that the upward departure was not justified under the facts of this case, we reverse and remand for resentencing.
FACTUAL HISTORY AND PROCEDURAL BACKGROUND
I. Facts of Current Offense
On January 30, 2003, Bad Marriage was released from tribal jail on the Blackfeet *1105Indian Reservation so that he could attend an Alcoholics Anonymous meeting. Instead, he went to the home of Leeta Old Chief, his girlfriend. After having consensual sex, the couple drove to visit friends. At Bad Marriage’s sister’s house, an argument ensued between Bad Marriage and Old Chief, and he began hitting her. The two went to the old rodeo grounds behind his sister’s home. There, Bad Marriage kicked and beat Old Chief in the thighs and pelvis.
At the rodeo grounds, Bad Marriage and Old Chief had anal sex. Old Chief later gave conflicting accounts of whether the sex was consensual or forced. She initially told an FBI agent, and Bad Marriage’s mother, that she had been raped. Bad Marriage was indicted for aggravated sexual abuse based on that allegation. Old Chief later retracted the claim, and said she had falsely accused Bad Marriage because she was angry at him for beating her and for corresponding with another woman.
Once Old Chief told law enforcement officers that she would no longer be willing to testify that Bad Marriage had forced her to have sex, the government dismissed the indictment in exchange for Bad Marriage’s agreement to plead guilty to assault resulting in serious bodily injury, in violation of 18 U.S.C. §§ 113(a)(6) and 1153. Bad Marriage accordingly pled guilty.
II. Bad Marriage’s Criminal and Social History
The Presentence Report prepared following Bad Marriage’s guilty plea documented 35 prior convictions from state courts in Washington and Montana. In addition, the report showed approximately 60 convictions in the Blackfeet Tribal Court. Bad Marriage’s adult state criminal record consisted of crimes committed between 1988, when the defendant was nineteen, and 2002. These convictions were almost entirely misdemeanors resulting in little or no jail time.
Four of his state court convictions were for assault: (1) a 1991 fourth degree as-saull/domestic violence conviction resulting in a ten day jail sentence; (2) a 1997 assault conviction resulting in a $120 fine; (3) a 1999 assault conviction for which he received a $270 fine; and (4) a 2002 conviction for partner/family member assault for which he served eight days.1
The record also included a 1989 second-degree burglary conviction for which Bad Marriage served thirty days in jail; a 1989 conviction for unlawful use of a weapon; a 2000 conviction for driving under the influence of alcohol (“DUI”); a 1993 conviction for violation of a contact order, resulting in his longest sentence of thirty-seven days; ten convictions for criminal trespass; over fifteen convictions for theft; and several convictions for obstructing a police officer or for disorderly conduct.
All of the theft incidents described in the Presentence Report appear to be shoplifting offenses. In each of these cases, Bad Marriage is described as having stolen beer, cigarettes, or chips from a convenience store. For example, in July 1999, Bad Marriage was arrested, and later convicted, three times for stealing a bag of chips or eating nachos at a convenience store. The most expensive item that Bad Marriage is described as having stolen is a few cases of beer.
*1106Most of Bad Marriage’s 60 convictions from Blackfeet Tribal Court were for disorderly conduct or public intoxication. Several convictions were for escape from tribal jail or for assault. These convictions, in accordance with U.S.S.G. § 4A1.2(i), were not counted for purposes of determining the defendant’s criminal history score.
Bad Marriage was intoxicated during numerous incidents described in the Pre-sentence Report. Of the 23 state criminal convictions for which any factual detail is provided, evidence of intoxication or alcohol use is apparent from seventeen of these convictions: these include incidents where Bad Marriage is described as intoxicated, where he attempted to steal alcohol, or where he engaged in disorderly conduct in a bar.
In addition, the Presentence Report describes Bad Marriage’s long history of alcohol and drug abuse. Bad Marriage first tried alcohol at the age of nine, and began regularly using it at eighteen. He last used alcohol on the night of the instant offense. Since his thirties, Bad Marriage also regularly smoked marijuana. Bad Marriage advised the probation officer preparing his Presentence Report that he was in need of treatment. The Report indicated that Bad Marriage was admitted to a treatment center in August 1988, but did not describe whether he completed that treatment. In 1998, Bad Marriage was admitted to the Blackfeet Chemical Dependency Center and dismissed one month later with a diagnosis of alcohol and cannabis dependence. That treatment center reported that he made-satisfactory progress during treatment, and that his prognosis was good provided that he followed his continuing care plan. No information is provided on any treatment or progress after that date.
The Presentence Report attributed one criminal history point each to seven of Bad Marriage’s convictions in accordance with U.S.S.G. § 4A1.1. Points were assigned for: three of the assault convictions;2 a 1999 criminal trespass and disorderly conduct offense; a 1999 conviction for theft and disorderly conduct; a 1999 conviction for theft; and the 2000 DUI conviction. No points were assigned to Bad Marriage’s shoplifting convictions because of this Court’s decision in United States v. Lopez-Pastrana, 244 F.3d 1025 (9th Cir.2001), which held that shoplifting offenses should be excluded from computation under U.S.S.G. § 4A1.2(c)(1).
Bad Marriage’s convictions resulted in seven criminal history points, but his total was capped at four points pursuant to U.S.S.G. § 4Al.l(c). This score established a criminal history category of level III.
III. Sentencing
At sentencing, the District Court ruled, based on U.S.S.G. § 4A1.3 (policy statement), that the criminal history computation of level III underrepresented both the seriousness of Bad Marriage’s past criminal conduct and the likelihood that he would commit future crimes. The Court accordingly departed upward one level to assign Bad Marriage a criminal history score of IV. The Court sentenced Bad Marriage to 41 months in prison, the high end of the sentencing range determined by the Guidelines for defendants with his offense score and adjusted criminal history level. In addition, the Court sentenced Bad Marriage to three years of supervised release, including the conditions that Bad *1107Marriage participate in a substance abuse program, abstain from the consumption of alcohol, and obtain mental health treatment.
During the hearing, the District Court explained its decision to depart upward as follows:
I look at this matter in its entirety. And I make the determination that I make based upon the entire history that is before me, taking into account all of the components of this history, which include 35 or more separate adult convictions, one of which was for the unlawful use of — with a weapon, two of which are for domestic or partner assault, a total of three by the record before me, involved one manner or another of as-saultive behavior.
The guidelines clearly contemplate that this court may take into account, not necessarily count in some numerical fashion, but take into account tribal convictions, which on the record before me are some 60 or more in number .:. Taking all of that into account, the record, whether it be counted for history points per se or not, is one of almost 100 separate convictions in this man’s adult life, substantial number of which have involved behavior directed to other human beings.
This is an individual who on the record has essentially an adult life devoted to criminal conduct, and much of it of a violent nature. And I conclude on the basis of this record that there is indeed every likelihood that this individual would return to criminal conduct in the future if given the opportunity. And it is therefore my finding that both components of 4A1.3 have been met on this record.
The day after the sentencing hearing, the District Court amended its statement of reasons to state that its determination to depart upward was made “without regard to tribal court convictions which, as urged by Defendant, expressly are not taken into account.”
Bad Marriage appealed his sentence.
DISCUSSION
I. Standard of Review
Because Bad Marriage was sentenced after April 30, 2003, the effective date of the Prosecutorial Remedies and Other Tools to End the Exploitation of Children Today (PROTECT) Act, the District Court’s decision to depart from the applicable sentencing guideline is reviewed de novo. United States v. Alfaro, 336 F.3d 876, 880 (9th Cir.2003). Previously, we reviewed a District Court’s departure decision for an abuse of discretion, as mandated by Koon v. United States, 518 U.S. 81, 98, 116 S.Ct. 2035, 135 L.Ed.2d 392 (1996). Id. As amended, 18 U.S.C. § 3742(e)(3)(B) provides that if a sentence is outside the applicable guideline range, the reviewing court must determine, inter alia, whether the departure is based on a factor that “does not advance the objectives set forth in section 3553(a)(2)” or “is not justified by the facts of the case.” The new standard still requires the Court of Appeals to “give due regard to the opportunity of the district court to judge the credibility of the witnesses” and to review the district court’s findings of fact only for clear error. 18 U.S.C. 3742(e).
II. Analysis
Bad Marriage contests the District Court’s decision to depart upward from the recommended U.S. Sentencing Guideline range. He contends that his criminal history, while extensive, does not consist of serious offenses warranting departure under U.S.S.G. § 4A1.3.
*1108A. Ninth Circuit Law on § 4A1.3 Departures
U.S.S.G. § 4A1.3 provides: “If reliable information indicates that the criminal history category does not adequately reflect the seriousness of the defendant’s past criminal conduct or the likelihood that the defendant will commit other crimes, the court may consider imposing a sentence departing from the otherwise applicable guideline range.” The Guidelines further note that departure is warranted only where the criminal history category “significantly under-represents” the defendant’s past conduct or likelihood of recidivism. Id.
As an initial matter, we note that following the Supreme Court’s decision in Blakely v. Washington, — U.S. --, 124 S.Ct. 2531, 159 L.Ed.2d 403 (2004), there is some question as to whether the facts supporting such a departure must be found by a jury beyond a reasonable doubt (or admitted by the defendant). Blakely held that a defendant’s Sixth Amendment right to trial by jury was violated when his sentence was increased based on a factual finding made by a judge rather than a jury. We recently ruled that although Blakely concerned Washington state law, its holding applied equally to the U.S. Sentencing Guidelines. United States v. Ameline, 376 F.3d 967 (9th Cir.2004). Although the Supreme Court has not required a court seeking to enhance a sentence to submit the fact of a prior conviction to a jury, see Apprendi v. New Jersey, 530 U.S. 466, 490, 120 S.Ct. 2348, 147 L.Ed.2d 435 (2000), it is possible that a § 4A1.3 departure still implicates Blakely where the enhancement is based on facts other than past convictions. Because we invalidate the departure in this case on other grounds, we do not resolve whether, or how, Blakely affects upward departures under § 4A1.3.
Under § 4A1.3, sentencing courts may take into account such information as:
(a) prior sentence(s) not used in computing the criminal history category (e.g., sentences for foreign and tribal offenses); (b) prior sentence(s) of substantially more than one year imposed as a result of independent crimes committed on different occasions; (c) prior similar misconduct established by a civil adjudication or by a failure to comply with an administrative order; (d) whether the defendant was pending trial or sentencing on another charge at the time of the instant offense; (e) prior similar adult criminal conduct not resulting in a criminal conviction.
Both the plain language of U.S.S.G. § 4A1.3, as well as Ninth Circuit precedent, establish that under that provision, a finding that either the seriousness of a defendant’s past conduct or the likelihood of recidivism was underrepresented can support an upward departure. United States v. Connelly, 156 F.3d 978, 984 (9th Cir.1998). In Connelly, the Court approved an upward departure where it found that the defendant’s past conduct was “serious,” but held that the departure was “also justified purely on the basis of Defendant’s likelihood of recidivism.” Id. at 984-85.
However, Connelly also left open the possibility that even though only one prong of U.S.S.G. § 4A1.3 must be satisfied, uncounted criminal conduct used to support an upward departure “must in all cases cross a threshold of ‘seriousness’.” Id. at 984. This issue remains unresolved. In United States v. Martin, we noted that “there is some uncertainty as to whether uncounted conduct must, as a threshold matter, be serious in order to be considered in deciding whether to depart.” 278 F.3d 988, 1002 n. 4 (9th Cir.2002). But we stated: “As in Connelly, because Defen*1109dant’s conduct was serious enough to meet such a requirement if it exists, we need not and do not resolve this uncertainty.” Id.
At least two Ninth Circuit cases have invalidated upward departures based on prior convictions that the Court found not to be serious. In United States v. Brady, this Court vacated an upward departure for a voluntary manslaughter defendant who had two prior tribal misdemeanor assault and battery convictions. 928 F.2d 844 (9th Cir.1991), overruled in part on other grounds by Nichols v. United States, 511 U.S. 738, 114 S.Ct. 1921, 128 L.Ed.2d 745 (1994), and by United States v. Watts, 519 U.S. 148, 117 S.Ct. 633, 136 L.Ed.2d 554 (1997). Although the District Court in that case had apparently relied only on the “recidivism prong” of U.S.S.G. § 4A1.3, Brady noted that the five examples listed under that provision in the Guidelines all refer to “serious” or “large scale” criminal conduct, suggesting that some level of seriousness is required. Id. at 853.
The Guidelines note:
Examples might include the case of a defendant who (1) had several previous foreign sentences for serious offenses, (2) had received a prior consolidated sentence of ten years for a series of serious assaults, (3) had a similar instance of large scale fraudulent misconduct established by an adjudication in a Securities and Exchange Commission enforcement proceeding, (4) committed the instant offense while on bail or pretrial release for another serious offense or (5) for appropriate reasons, such as cooperation in the prosecution of other defendants, had previously received an extremely lenient sentence for a serious offense.
(emphasis added)
We stated that, in contrast to those examples, “Neither of Brady’s tribal assault convictions, both sentences of less than 30 days, represent this level of seriousness.” 928 F.2d at 853. Although Brady also went on to hold that a defendant’s sentence could not be enhanced pursuant to uncounseled tribal convictions, id. at 853-54, the fact that Brady’s assault convictions were not serious was an independent basis — indeed the first reason — for our decision.
Likewise, United States v. Carrillo-Alvarez found that a defendant’s uncounted past conviction for auto burglary was neither “serious” nor “large scale,” and did not sufficiently distinguish his conduct from that of other offenders assigned the same criminal history level. 3 F.3d 316, 322 (9th Cir.1993).
In contrast to Brady and Carrillo-Alvarez, other Ninth Circuit cases have approved upward departures without inquiring into the seriousness of the defendant’s past crimes. For example, we have permitted upward departures based primarily on the similarity of the defendant’s prior conduct to the instant offense, recognizing that such similarity “entails greater culpability” and reflects on a likelihood of recidivism. United States v. Segura-Del Real, 83 F.3d 275, 277 (9th Cir.1996) (internal citation omitted) (upholding increased sentence for individual convicted of illegal reentry to the United States where he had several prior immigration violations, even though District Court characterized those offenses as minor). See also United States v. Goshea, 94 F.3d 1361, 1364 (9th Cir.1996) (sustaining departure based on prior similar conduct in impersonating military officer).
In addition, while the examples listed under U.S.S.G. § 4A1.3 all refer to “serious” or “large scale” crimes, another note in the Guidelines suggests that in at least some cases, the sentencing court may consider non-serious prior convictions. The Guidelines establish a time span (either ten or fifteen years, depending on the *1110crime) in which prior criminal convictions are assigned criminal history points. See U.S.S.G. § 4A1.2 (e). While convictions obtained before the applicable time period are not assigned points, if a court finds that they are “evidence of similar, or serious dissimilar, criminal conduct, the court may consider this information in determining whether an upward departure is warranted under § 4A1.3 ....” U.S.S.G. § 4A1.2 n. 8 (emphasis added). This suggests that while dissimilar criminal conduct committed before the relevant time period must be serious in order to be considered, remote convictions that are similar to the instant offense may be considered even if they are not serious.3
While the Guidelines and Ninth Circuit precedent do not establish that prior convictions must always be serious in order to support an upward departure, they certainly indicate that the seriousness of past convictions is an important factor to be considered. Our decisions in Brady and Carrillo-Alvarez, in addition to the examples provided by the Guidelines under § 4A1.3, support that conclusion.4 The importance of the seriousness of prior convictions is also underscored by our cases emphasizing that “it is the quality of the defendant’s criminal history not the quantity which is decisive.” Segura-Del Real, 83 F.3d at 277 (citing Carrillo-Alvarez, 3 F.3d at 322-23). “An upward departure under § 4A1.3 may not be based on mechanical calculations. It must be justified, rather, by a qualitative difference between the defendant’s overall record and that of other defendants in the same criminal history category.” Carrillo-Alvarez, 3 F.3d at 322.
These “quality over quantity” statements were made in cases involving defendants assigned a criminal history level VI, the highest level, and the rationale was based in part on the fact that all level VI defendants have lengthy criminal records. Thus, these courts reasoned that one level VI defendant could not be distinguished from another solely on the basis of a higher criminal history score. See Carrillo-Alvarez, 3 F.3d at 323. Since Bad Marriage was only assigned to level III, which includes defendants with fewer convictions than those assigned to level VI, the dissent believes that Bad Marriage’s likelihood of recidivism supports an upward departure under U.S.S.G. § 4A1.3. But the dissent fails to realize that, despite the particular context in which these “quality over quantity” statements were made, the principle is more generally applicable: the sheer length of a defendant’s criminal history should not, on its own, support a departure under U.S.S.G. § 4A1.3.
[5] Therefore, we hold that the seriousness of a defendant’s prior convictions must be a significant factor in a decision to depart under either prong of § 4A1.3.
*1111B. “Seriousness Prong”
We now examine whether the upward departure was justified according to the facts of this case. We first consider whether a departure was appropriate under the first prong of § 4A1.3, that “the criminal history category does not adequately reflect the seriousness of the defendant’s past criminal conduct.” We bear in mind that an upward departure “is appropriate only in an unusual case, because the criminal history category of the sentencing guidelines is designed expressly to account for a defendant’s prior criminal conduct.” Carrillo-Alvarez, 3 F.3d at 320 (citing United States v. Singleton, 917 F.2d 411, 412 (9th Cir.1990)).
Here, the sentencing judge based his upward departure in part on Bad Marriage’s previous convictions for assault. The judge stated: “I look at this matter in its entirety. And I make the determination that I make based upon the entire history that is before me, taking into account all of the components of this history, which include 35 or more separate adult convictions, one of which was for the unlawful use of — with a weapon, two of which are for domestic or partner assault, a total of three by the record before me, involved one manner or another of assaultive behavior.”
However, Bad Marriage’s prior convictions that have already been assigned points, including three of his four assault convictions, cannot be used to support a departure under § 4A1.3. In United States v. Henderson, 993 F.2d 187, 189 (9th Cir.1993), this Court held that a sentencing court could not depart upward based on the nature of a defendant’s crimes already assigned criminal history points. “A factor already calculated into a sentence under the guidelines may not be a proper basis for departure.” Id. at 189 (quoting United States v. Hernandez-Vasquez, 884 F.2d 1314, 1315 (9th Cir.1989));
The government attempts to distinguish Henderson on the grounds that Bad Marriage’s ■ criminal history score was capped at four points, pursuant to § 4Al.l(c), even though he actually earned seven points. Therefore, it contends that Bad Marriage’s scored convictions can be considered again under § 4A1.3. We disagree. U.S.S.G. § 4A1.1 establishes a system for assigning criminal history points according to the severity of prior convictions. Three points are assigned for each prior sentence exceeding one year and one month; two points are assigned for each prior sentence of at least 60 days, but less than one year and one month; and one point for each prior sentence of less than 60 days. While point totals in the first two categories are not,capped, the Guidelines specifically limit the number of points that can be computed for prior sentences less than 60 days. The apparent reason for this distinction is that the Guidelines consider a lengthy record of short sentences to be less indicative of culpability, the potential for recidivism, or other characteristics warranting greater punishment. If a sentencing court could depart upward based on sentences assigned criminal history points, but excluded pursuant to § 4Al.l(c), it would defeat the Guidelines’ evident intention to limit the number of minor past convictions bearing on a sentencing determination.
Thus, the District Court erred in double-counting several of Bad Marriage’s prior convictions. We remove from consideration those convictions that were already scored, including three of the four assault convictions; convictions for criminal trespass and disorderly conduct; theft and disorderly conduct; theft; and DUI.
Most of the remaining convictions that we may properly consider are for disorderly conduct, criminal trespass, obstructing a police officer, unlawful use of a weapon, or *1112theft. These are all minor offenses. For instance, all of the theft convictions in the Presentence Report for which any detail is provided are for shoplifting items of trifling value (nachos, cigarettes, beer, or potato chips), usually when Bad Marriage was drunk. Certainly, evaluated in light of their “quality rather than quantity,” these offenses do not make Bad Marriage’s case the “unusual case” warranting departure.
The most significant of Bad Marriage’s uncounted prior convictions are his 1991 conviction for fourth degree assault, a 1989 conviction for second-degree burglary, and a 1993 conviction for violation of a no-contact order. None of these convictions makes Bad Marriage’s criminal history significantly more serious than that of other level III defendants.
The 1991 assault conviction was for fourth degree assault/domestic violence and resulted in a ten day sentence. This conviction is similar to the assault sentences found inadequate to support a § 4A1.3 departure in Brady. Like Brady, Bad Marriage was sentenced to less than thirty days in prison. Although we found a defendant’s prior assault with a deadly weapon serious enough to justify an upward departure in United States v. G.L., 143 F.3d 1249, 1255 (9th Cir.1998), the class of assault that Bad Marriage was convicted of, and the short sentence imposed, suggest a less serious offense. The Presentence Report contained no additional factual detail on the offense, making it impossible to consider it serious on the basis of the underlying conduct. See United States v. Wyne, 41 F.3d 1405, 1408 (10th Cir.1994) (deeming assault conviction insufficiently “serious” based on limited information provided, while noting that this determination did not imply that assault on a female was not serious in the common sense of the word).
The second-degree burglary conviction also does not seem to be serious in light of the thirty-day jail sentence imposed and in comparison with Carrillo-Alvarez, where the court found a conviction for auto burglary insufficiently serious to depart upward. 3 F.3d at 322. Similarly, judging by the limited information available, Bad Marriage’s uncounted conviction for violation of a no-contact order, for which he served his longest prison term of thirty-seven days, does not make his criminal history so unusual as to distinguish him from other defendants in his category.
The government contends that even if Bad Marriage’s convictions do not appear serious when viewed individually, their sheer quantity and repetitiveness justify an upward departure. The District Court stated: “[T]he record ... is one of almost a hundred separate convictions in this man’s adult life, substantial number of which have involved behavior directed to other human beings. This is an individual who on the record has essentially an adult life devoted to criminal conduct, and much of it of a violent nature.”
That determination ignored several facts. First, as noted, most of Bad Marriage’s prior offenses involving violence were already scored. Second, Bad Marriage’s criminal record does not show either an attempt to make a living off of crime or an escalation in severity. See Singleton, 917 F.2d at 413 (sustaining upward departure for a defendant with a “pattern of consistent criminal conduct that had started with criminal mischief and had escalated into serious felonies”). Bad Marriage’s record reveals an individual ravaged by substance abuse, not a depraved criminal. We do not hold that a series of misdemeanors can never support an upward departure under the first prong of § 4A1.3, but we conclude that a depar*1113ture was not justified under the facts of this case.
C. “Recidivism Prong”
We further hold that an upward departure was not justified according to the second prong of § 4A1.3, based on “the likelihood that the defendant will commit other crimes.” Connelly identified three factors that sentencing courts should consider in evaluating the potential for recidivism: “1) the quantity (or ‘repetitiveness’) of uncounted criminal conduct, 2) the similarity of uncounted criminal conduct to the offense conduct, and 3) the degree to which the defendant has been deterred by prior sentences.” 156 F.3d at 985. Without considering the seriousness of Bad Marriage’s prior convictions, the dissent believes that all three prongs weigh in favor of an upward departure because Bad Marriage has 1) received thirty-five state criminal convictions, 2) most of which involved alcohol, and 3) none of which had any deterrent effect. But it is clear from Brady and Carrillo-Alvarez, as discussed above, that the seriousness of prior convictions is an equally important factor in evaluating the likelihood of recidivism.
Applied to Bad Marriage, the first factor, the quantity of uncounted criminal conduct, suggests a potential for recidivism. The other factors, however, pull in the opposite direction. Only one of Bad Marriage’s uncounted criminal convictions — the 1991 assault/domestic violence conviction, for which he served ten days in jail — appears to be similar to the current offense. By far, the bulk of Bad Marriage’s criminal history consists of minor offenses, involving no threat to human beings, committed while the defendant was intoxicated.
With regard to the third factor, the degree to which the defendant has been deterred by prior sentences, we note that Bad Marriage has never served in prison for longer than thirty-seven days at a stretch. Without any enhancement, the Sentencing Guidelines set a range of 27 to 33 months for a defendant with Bad Marriage’s offense and criminal history level. Even the most lenient sentence within this range is more than twenty times the length of his longest prior period of incarceration. Although Bad Marriage has had numerous prior convictions resulting in short sentences, nothing in his record suggests that the standard range set by the Sentencing Guidelines would be an inadequate deterrent.
Finally, we have already analyzed above the seriousness of Bad Marriage’s prior criminal record and found it to be far less serious than the examples listed in the Sentencing Guidelines as warranting departure. We are certain that the Sentencing Commission, in providing for upward departures based on an exceptional likelihood of recidivism, was not contemplating an individual who might steal more nachos from a convenience store or return to swearing in public while drunk. We do not in any way minimize the gravity of Bad Marriage’s assault on Old Chief or his prior acts of assault. However, we believe that his criminal history calculation, which included three points assigned for his prior assaults, adequately captured that history. Considering all four factors together, we conclude that an upward departure based on the second prong of § 4A1.3 was not justified.
D. Substance Abuse
Our evaluation of this case is colored by the patent role of substance abuse in creating the tragedy before us. No one reading Bad Marriage’s Presentence Report can fail to be struck by the evidence of alcohol abuse running through it. Although alcohol abuse cannot excuse criminal behavior, we cannot ignore the presence of this fac*1114tor in determining the appropriateness of an enhanced sentence.
The First Circuit has noted that when reviewing whether a departure is “justified by the facts of the ease,” a court of appeals may consider the underlying purposes of sentencing, the Sentencing Commission’s intent, and other extra-record sources. United States v. Thurston, 358 F.3d 51, 76 (1st Cir.2004). The standard of review “calls for an evaluative judgment, not a mechanical exercise.” Id.
In our evaluative judgment, an upward departure is inappropriate where a defendant’s criminal past is composed of mostly minor crimes committed under the influence of alcohol. The underlying purposes of sentencing include not only punishment and deterrence, but also the provision of treatment to a defendant in need of it. 18 U.S.C. § 3553(a)(2)(D).
In United States v. Luscier, 983 F.2d 1507, 1510-11 (9th Cir.1993), we invalidated an upward departure that was based on a defendant’s prior drug abuse. We pointed out that the Sentencing Guidelines recognize that drug abuse may be correlated with recidivism but do not list it as a factor warranting upward departure. Id. at 1510. Instead, the Guidelines recommend that an offender with a history of drug or alcohol abuse be required to participate in a rehabilitation program as a condition of supervised release. Id. (citing U.S.S.G. § 5H1.4 policy statement).
It appears that, faced with a choice between competing policies, the Commission opted for a sentencing scheme that encourages defendants to admit to and seek treatment for drug dependency, rather than one that treats more severely defendants who have a history of drug abuse. Thus, while in the ordinary case drug addiction will influence the conditions of supervised release, past drug use should not affect time in custody unless the defendant’s condition is so extraordinary that departure, rather than the measures discussed in section 5H1.4, is required.
Id.
In this ease, the length and character of Bad Marriage’s criminal record is clearly the result of a serious drinking problem. To sentence Bad Marriage to a longer prison term on the basis of that record would serve no useful purpose.
Alcohol abuse on Indian reservations is a social problem of devastating scope.5 Almost twenty years ago, Congress passed the Indian Alcohol and Substance Abuse Prevention Treatment Act of 1986, 25 U.S.C. § 2411, recognizing that alcohol and substance abuse presented the most severe health and social problem facing Indian tribes and individuals. See 25 U.S.C. § 2401 (listing Congressional findings supporting that legislation). Congress expressed concern that Indians die from alcoholism at a rate four times that of the U.S. population as a whole, and that substance abuse results in a rate of years of potential life lost almost five times that of the United States at large. Id. at § 2401(5).
Despite increasing recognition of the problem, the consequences of substance abuse on Native Americans are still staggering. The U.S. Department of Health and Human Services reports that American Indian men age 25 to 34 are seven times more likely to suffer from health problems related to alcohol, such as cirrhosis of the liver, than non-Indian men in the same age group; they are twice as likely to commit suicide; and they are almost three times as likely to die in motor vehicle *1115crashes.6 U.S. Department of Health and Human Services, Prevention Alert: American Indians/Alaskan Natives and Substance Abuse (2002), http://neadi.samh-sa.gov/govpubs/prevalert/v5/10.aspx. While many Native Americans drink less frequently than members of other ethnic groups, the problem appears to be concentrated among young people and a segment of drinkers over 26 who are heavy alcohol users. Id.
Alcohol abuse shapes Native Americans’ encounters with the criminal justice system. The rate of American Indians arrested for alcohol violations (DUI, liquor law violations, and public drunkenness) in 1996 was more than double that of the general population. U.S. Department of Justice Bureau of Justice Statistics, American Indians and Crime 25 (1999). In addition, about 70 percent of American Indians in local jails nationwide convicted of violent offenses reported that they had been drinking when they committed the offense. Id. at 29. Three out of four American Indian victims of family violence reported that they perceived the perpetrator to have been drinking at the time of the incident, compared to half of the victims of other ethnic and racial groups. Id. at 9-10 (summarizing data from 1992-1996).
Native Americans who come into contact with the criminal justice system are more likely to be charged, and sentenced, under federal law than the average American. Because tribal courts have only limited jurisdiction in criminal matters, serious offenses are charged under U.S. law. As a result, Native Americans are disproportionately sentenced under the U.S. Sentencing Guidelines. Federal courts must be keenly aware of the underlying social problems facing the Native American offenders who come before them, such as alcohol abuse, and of the need of many of these defendants for rehabilitation.
Treatment is critical. “Without treatment during confinement, prisoners with substance abuse problems cannot recover from their addictions and have no viable way to prepare for their return to society where they might face the same environmental risk factors that triggered their substance abuse.” U.S. Department of Justice Office of Justice Programs, Promising Practices and Strategies to Reduce Alcohol and Substance Abuse among American Indians and Alaskan Natives 35 (2000). New research indicates that successful interventions among Native American drinkers often build upon tribal culture and beliefs. Id. at ix. See also id. at 1-38 (describing culturally reflective Native American substance abuse treatment programs).
While the dissent correctly states that this opinion will not alleviate alcohol abuse on Indian reservations, it is clear that prison alone will neither rehabilitate an individual critically in need of substance abuse treatment, nor, in the long run, protect society against him. Bad Marriage is such an individual.
Here, an upward departure pursuant to § 4A1.3 was not justified by the facts of the case.
E. Conclusion
We hold that the upward departure in this case was not justified by the facts, and remand to the District Court for resen-tencing within the appropriate range.
REVERSED AND REMANDED.
. Because the U.S. Sentencing Guidelines define a "sentence of imprisonment” so as to exclude a suspended portion of a sentence, see U.S.S.G. § 4A1.2(b)(2), we also list only the non-suspended portions of sentences. All citations to the U.S. Sentencing Guidelines in this opinion are to the version incorporating amendments effective November 1, 2002, except where otherwise indicated.
. The first assault conviction was not counted under U.S.S.G. § 4A1.2(e)(3) because it was more than ten years old.
. Although Brady may be read as precluding sentencing courts from considering non-serious past convictions even if they are similar to the current offense, that case was decided in 1989, before the Sentencing Guidelines were amended to include the language “or serious dissimilar.’’ See United States v. Smallwood, 35 F.3d 414, 418 n. 8 (9th Cir.1994).
. The dissent contends, however, that we have ignored the plain language of U.S.S.G. § 4A1.3 and "instead relie[d] on several past cases and other sections of the Guidelines to rewrite [the statute] in a misplaced effort to right what [we] perceive as a social wrong.” The dissent is correct in part of its criticism: we have relied on "several past cases” to adjudicate this case. Indeed, to decide like cases alike and prevent intracircuit conflict, panels of this court are obligated to follow precedent as we have done here. See United States v. Frank, 36 F.3d 898, 902 (9th Cir.1994). Thus, the majority is not rewriting, but rather interpreting, the statute by relying on this Court's precedent.
. In this opinion, we use both "American Indian” and "Native American,” as both terms are used by the community. When citing statistics, we adopt the usage of the original source.
. At the time of the offense, Bad Marriage was 33 years old.