« Big day for guns (and other fun?) at SCOTUS | Main | Prosecutors urged lighter sentence in 2000 for sex offender suspected in missing California teen case »

March 2, 2010

SCOTUS further restricts the reach of ACCA mandatory minimums in Johnson

Justice Antonin Scalia has proved once again that he is the friend of federal criminal defendants through his opinion for the Court today ruling in such a defendant's favor in Johnson v. United States.  Here is a brief account of the ruling via SCOTUSblog:

Johnson v. United States (08-6925), the Court rules 7-2 that a “violent felony” under federal law requires the use of physical violence, thereby reversing and remanding the lower court.  Justice Scalia writes for the majority, while Justice Alito dissents, joined by Justice Thomas.  The full opinion in pdf format is here.

Because I am on the road and then teach this afternoon, I may not have a chance to fully process and comments on this ruling until late tonight. In the meantime, I hope informed readers might comment on whether they think Johnson is a big deal or just another little (pro-defendant) tweak of the federal Armed Career Criminal Act.

March 2, 2010 at 11:29 AM | Permalink


TrackBack URL for this entry:

Listed below are links to weblogs that reference SCOTUS further restricts the reach of ACCA mandatory minimums in Johnson:


tweak. a correct one, but just a tweak.

"the friend of federal criminal defendants" really, doug, are you letting supremacy claus write your openings?

Posted by: big bad wolf | Mar 2, 2010 1:07:25 PM

I am the mother of a young man serving a 5 year minimum sentence for "distribution of pornography to a minor" this happened as a result of a Federal Internet Sting. Because this offense involved a "minor" and he had less than 100 pictures of what was considered child pornography (this was deleted but forensic experts recovered the images) he is considered a "Violent Sex Offender" and will be labeled as such on the Registered Sex Offender Resistry. How can this be when there was no violence or force involved? Our Attorney said because anything involving a child is considered violent. Will this ruling help others in my son's situation to remove the violent label?

Posted by: Miriam Young | Mar 2, 2010 1:34:28 PM

Miriam: The ability to do something depends upon when the issue is raised in the proceedings. The best opporunity, raising it in a sentencing hearing, appears to be lapsed. If the case is on appeal, it might be possible to raise it in that context, although it may be difficult to do so if this was not raised in the trial court.

In a collateral attack (i.e. after conviction and a direct appeal), via either a post-conviction motion or a habeas corpus petition, there are intricate procedural issues involved. One must first determine if the ruling has retroactive effect; one must also determine if the time to assert a right has been exhausted. This is a case that it might be worth investigating with someone who regularly does post-conviction work, and promptly if it is to be done at all.

Posted by: ohwilleke | Mar 2, 2010 2:09:11 PM

I'm inclined to think that this is more than a tweak, because the methodology it embraces, of looking beyond the elements of the offense to the underlying conduct when the offense could include some conduct that does not constitute "physical violence" has wide applicability, and because "physical violence" is a higher standard than the "physical force" standard that was being applied previously. I think that there are lots of cases out there that could fall into the "Johnson gap" where there are predicate convictions that could include conduct that involves physical violence, but didn't necessarily actually include such conduct.

The biggest question mark the Johnson case leaves, it seems to me, is how one handles nolo contendre or other factually ambiguous pleas of guilty to a Johnson gap offense.

Posted by: ohwilleke | Mar 2, 2010 2:13:49 PM

ohwilleke, I think you read the opinion wrong.

The opinion does not allow one to look at the "underlying conduct" in determining if a prior offense is a violent felony. The opinion only says that sometimes (as in Floraida Battery) a state offense will have alternate ways of committing an offense. In other words, alternate elements of the statute. Some of those alternate ways will have, as an element, the use of physical force (as Johnson defined the term); and others will not have, as an element, the use of physical force. In that situation, you can look to reliable (Shepard-approved) documents to determine which set of alternative elements formed the basis for the prior conviction. As stated by the majority: you look to the documents to determine "which statutory phrase was the basis for the conviction." (Slip. Op. at 10).

Posted by: DEJ | Mar 2, 2010 2:26:01 PM

Here is a summary that I've written for other purposes and I will re-post here:

The Court (in an opinion written by Scalia) held 7-2 in favor of the Defendant’s argument: “[T]he phrase ‘physical force’ means violent force –that is, force capable of causing physical pain or injury to another person.” (Slip Op. at 6 (emphasis original)). Specifically, the Court held that Florida battery, which can be committed by “any intentional physical contact ‘no matter how slight’” (Slip Op. at 4 (emphasis original)) is NOT a “violent felony” under the ACCA. Alito and Thomas dissented.

So, what are the implications of this big win? There are many:

First, the Court makes clear that federal courts are bound by how a state court interprets the elements of a state statute. Accordingly, when the Florida Court says any slight touch may constitute a battery, the federal courts must accept that conclusion. This does not mean, however, that a federal court is bound by how a state courts interprets the phrase “physical force” in its own recidivism statute. (Slip Op. at 4-5). The interpretation of a federal statute’s use of the phrase “physical force” is an issue of federal law.

Second, the Supreme Court’s opinion should have broad application. The Court’s rationale should apply any time “physical force” is used to define a phrase that uses the term “violent” or “violence,” and is used to define a prior felony conviction. This means that, in addition to the ACCA (18 USC 924(e)), Johnson should also impact how federal courts interpret the phrase felony “crime of violence,” which is found in 18 USC 16(a), USSG 2L1.2 and USSG 4B1.2(a)(1).

Third, the Court explicitly says that it is not interpreting the phrase “physical force” under 18 USC 922(g)(9), which prohibits a person with a prior “misdemeanor crimes of domestic violence” from possessing a firearm. It stresses: “We do not decide that the phrase has the same meaning in the context of defining a misdemeanor crime of domestic violence. The issue is not before us, so we do not decide it.” (Slip Op. at 9-10). But see, US v. Hays, 526 F.3d 674 (10th Cir. 2007) (interpreting “physical force” in the same manner the Supreme Court did in Johnson).

The Court’s dicta in Johnson implies that it might view prior misdemeanor convictions to be different. Put another way, under the Court’s rationale in Johnson, the Supreme could conclude down the road that “physical force” has a different meaning for the purpose of defining a qualifying prior conviction under 922(g)(9) than it does under 924(e). In my view, however, this conclusion is not inevitable, as the Court did not consider the many arguments that support a narrow interpretation of 922(g)(9) as well.

Finally, Johnson is important because it expounds upon the “categorical” and “modified categorical” approaches. When there are no reliable documents to show what part of a divisible statute a person was previously convicted under, then the federal court is to assume it was for “the least” serious conduct. (Slip Op. at 2). However, if the record does include such documents, then you can use the modified approach “to determine which statutory phrase was the basis for the [prior] conviction.” (Slip Op. at 10). If the approved documents, however, do not show what part/phrase of the statute formed the basis for the prior conviction, then it cannot be classified as a “violent felony” or “crime of violence.”

This language makes clear that a look to Shepard-approved documents is NOT used to determine what the defendant in fact (or in actuality) did. Again, the modified categorical approach is only used to determine "which statutory phrase was the basis for the conviction." Further, when you are talking about “force” as an “element” of a prior offense, reliance on Shepard-approved documents is only to be used when the statute itself is divided into different “statutory phrase[s]” or parts.

Ultimately, Johnson is a great opinion that can be very helpful when arguing that a prior felony conviction is not a “violent felony” or “crime of violence.”

Posted by: DEJ | Mar 2, 2010 2:31:44 PM

DEJ, I don't think I meant to read the opinion very much differently than you are suggesting, although that may not have been clear. I certainly don't disagree that Johnson is irrelevant to offenses where physical violence is an element of the offense (aggravated assault), and where physical violence is irrelevant to establishing that an offense has been committed (larceny or embezzlement), and I certainly don't disagree that resolution of the issue through Shepard approved reliable documents is conclusive when they resolve the issue.

It seems to me though, that there have to be lots of offenses, like the battery offense in Johnson, in which the same phase, e.g., "involving any physical contact," could include offenses that both are and are not physically violent. Rather than being a tiny subset of all convictions, I suspect that Johnson offenses are a rather significant subset of all convictions.

At any rate, whether there is one phrase that includes cases that involve offenses involving and lacking physical violence as defined in Johnson, or two alternate phrases, one qualifying and one not, in the offense of conviction, I think it is safe to assume that it will be of no avail to look to "reliable (Shepard-approved) documents to determine which set of alternative elements formed the basis for the prior conviction. As stated by the majority: you look to the documents to determine 'which statutory phrase was the basis for the conviction.'"

If people in the criminal justice system weren't attuned to the Johnson distinction, presumably they often won't have made the distinction clear in the documentation of the case. I expect that law enforcement agencies, prosecutors and judicial bodies across the nation will be adding check boxes to their forms to clarify that distinction over the next year or two, but old documents will often be opaque, and old convictions will matter for a long, long time, because the applicability of the statute is to someone's criminal record which could have arisen long before the current conviction.

Suppose that a review of Shepard-approved documents provides no insight on the alternative that formed a basis of conviction in Johnson's case. Does he win, or do they have an evidentiary hearing? I did read the opinion (admittedly in a glance this morning) to permit evidence on this issue but it is perfectly possible that I am mistaken.

Posted by: ohwilleke | Mar 2, 2010 3:01:24 PM

Correction: "I think it is safe to assume that it will OFTEN be of no avail to look to . . ."

Posted by: ohwilleke | Mar 2, 2010 3:03:21 PM

ohwilleke, if a prior offense's element be committed by either violent acts or non-violent acts (e.g. "actually and intentionally touching another person"), then it cannot be counted as a violent felony because it does not have "as an element" the use of physical force. Period. End of story. That, in fact, is the holding of Johnson.

The only thing we knew about Mr. Johnson's prior conviction was that he committed the "least" of the statute's prohibitions. He "actually and intentionally touched" another person, in violation of FL's battery statute. Of course, this could be comitted via violence but it could also be comitted without violence. As such, it does not have the use of force "as an element."

You ask: "Suppose that a review of Shepard-approved documents provides no insight on the alternative that formed a basis of conviction in Johnson's case. Does he win, or do they have an evidentiary hearing?" This was the facts of Johnson, and the answer is, he wins. You do not have an evidentiary hearing to determine what actually happened to result in a prior conviction.

I don't deny you that a prior conviction's documents will sometimes not distinguish between a statute's various parts or phrases. Based on experience, I can tell you that sometimes they do, and sometimes they don't. But the majority also recognized this problem: "absence of records will often frustrate the application of the modified categorical approach." (Slip Op. at 11). Nevertheless, in order to prevent mini-trials on what a person was previously convicted of, see Taylor, this is the best course of action.

Posted by: DEJ | Mar 2, 2010 3:34:02 PM

I don’t think Justice Scalia is any “friend of federal criminal defendants.” He has ruled against them on plenty of occasions. In this case, he simply interpreted the law as he believed his oath of office requires of him. Scalia is by no means pro-defendant, but in certain types of cases his philosophy happens to point him in a certain direction.

Posted by: Marc Shepherd | Mar 2, 2010 4:14:33 PM

DEJ: I see it the same way - just a refinement of the old "modified categorical approach" if you will -- if a statute can be violated in ways that both include and exclude force then a court can look to the charging docs (etc) to see if they can pin down what part was charged -- if not -- lucky break in Cell Block C.

Posted by: Ferris Bueller | Mar 2, 2010 5:05:03 PM

DEJ's reading is the correct one, as James v. United States and Chambers v. United States made clear.

Looking at Shepard sources is only appropriate where a statute covers more than one category of behavior but it is unclear from the conviction which category was at issue. For example, the same statute could state that it is a crime of criminal escape to (1) break out of a confinement, or (2) failing to report to a probation officer. In this case, the Shepard document could be consulted to determine which of the two parts of the statute the defendant's conduct fell under, since meaningfully different behavior underlie each of the statutory components.

This is in contrast to a statute that does not contain multiple categories, such as Florida's battery statute, which contains a single category. In that case, you use the Begay test to determine if a statute was purposeful, violent, and aggressive based upon how the offense is ordinarily committed, as established by the elements of the offense. You cannot use Shepard sources to look to the underlying conduct, because there is only one category of conduct at issue.

Posted by: Res ipsa | Mar 2, 2010 5:43:56 PM

I'm a law student who clerked for several months at the federal defender office in my district. I worked quite a bit on the ACCA "violent felony" question, and I think the most important aspect of Johnson is its clarification of the modified categorical approach. Not that it completely clarifies the issue, but it provides some much-needed guidance to judges who misunderstand the purpose of the categorical approach and the mandates of Sixth Amendment in a post-Apprendi world.

The modified categorical approach allows a sentencing court to go behind the actual words of a statute. It can look at the various documents allowed by Shepard and Taylor, which will often reveal facts about the conduct leading to a defendant's prior convictions. Even so, the court cannot sentence the defendant based on those facts. The only two facts the court may consider are 1) that the defendant was convicted, and 2) that the crime, as defined by statute, fits the definition of a violent felony. The modified categorical approach exists solely to aid the courts in determining whether 2) is the case.

What this means is that a defendant may receive an enhanced sentence under the ACCA for certain types of crimes, but not for the conduct underlying those crimes. It is the crime that is violent, for ACCA purposes, and not the behavior. This is why the modified categorical approach is available only when the strict categorical approach does not provide a conclusive answer.

None of this is revolutionary, either. Almendarez-Torres says that the fact of a prior conviction is a sentencing factor (as opposed to a offense element), and therefore need not be proven 1) to a jury, or 2) beyond a reasonable doubt. Apprendi and Blakely make clear that any other fact which increases a defendant's potential sentence is an element that must be proven to a jury beyond a reasonable doubt. The modified categorical approach would therefore violate the Sixth Amendment if it allowed for greater sentences based on the facts underlying prior offenses.

Here is a good way to think about it. The modified categorical approach can ask one of two questions. Question One: Was the crime the defendant committed a violent felony? Question Two: Did the defendant, in committing the crime, act violently? The first question is the only one permitted by the Constitution. Under it, the facts are relevant only insofar as they help identify the crime committed by the defendant. Under the second question the facts are everything.

Unfortunately, courts often ask the second question. This is mostly due to the Court's fairly vague jurisprudence. The Court has explained the functioning of modified categorical approach, but it has never clearly spoken on its availability. The soundest approach is to limit it to cases where 1) the statute sets out different categories of prohibited conduct, and 2) it is not clear what category covers the defendant's conduct. The Seventh Circuit has taken this approach, see U.S. v. Woods, 576 F.3d 400, with the requirement that the various categories be clearly demarcated so the statute can be read to create multiple crimes, some violent and some not. I believe doing that is the best approach because it limits the temptation to sentence based on the facts. That, or overturn Almendarez-Torres.

ohwilleke: I think you are correct that most statutes are like the one in Johnson. Rather than being a problem, though, this fact is a good thing for defendants. Most statutes that are capable of being violated in both violent and nonviolent ways are typically committed in the nonviolent way. And as James v. U.S. tells us, a court may only look to the typical case. Broadly worded statutes are good for defendants.

Posted by: afm | Mar 2, 2010 9:34:33 PM

Marc Shepherd --

"I don’t think Justice Scalia is any 'friend of federal criminal defendants.' He has ruled against them on plenty of occasions. In this case, he simply interpreted the law as he believed his oath of office requires of him. Scalia is by no means pro-defendant, but in certain types of cases his philosophy happens to point him in a certain direction."


The suggestion that Scalia would decide a case to be friendly (or unfriendly) to a certain class of litigants is very misguided and demeans Scalia. It would be grossly improper for a judge to decide a case based his sympathies for the class to which a particular litigant belongs. Scalia has never done this.

For the same reason, Prof. Liu, nominated for by Obama the Ninth Circuit, was wildly off-base in criticizing Alito on the basis of Alito's "ruling for the state," and against largely poor or minority defendants, in Third Circuit death penalty cases. Since when are poor and minority litigants presumptively favored in law suits?

For Liu to criticize Alito or anyone else for decisions "favoring" this group or that shows such a warped view of what judging is about as to be per se disqualifying. Result-orientation used to be a dirty word, but for Prof. Liu, it is, if not the whole ball of wax, a big part of it.

Whites and well-off people get due process too, even if they voted predominantly for McCain. This Liu nomination is a disgrace, and will push the already outlier Ninth Circuit even more toward la-la land.

Posted by: Bill Otis | Mar 2, 2010 11:18:50 PM

afm--Very good explanation. A similar point was recently made in a Sixth Circuit opinion that overruled a prior Sixth Circuit precedent for doing precisely what you said they could not (the prior opinion was decided before James/Begay/Chambers). From what I know, the case is now being re-heard en banc.


I have to say, despite being a district judge sitting by designation, Judge Sara Lioi really nailed it.

Posted by: Res ipsa | Mar 3, 2010 9:04:25 AM

Thanks, Res Ipsa. I'll check it out.
The Fourth Circuit issued a decision last week that sort of supports my view, as well (http://pacer.ca4.uscourts.gov/dailyopinions/opinion.pdf/094336.P.pdf).

The Seventh Circuit is still the only circuit to do (what I consider) a proper analysis of the purpose of the categorical and modified approached. I cited it in my original comment. It's definitely worth a read. The interesting thing about the Woods case is that it's a Guidelines case discussing the modified categorical approach in the career offender guideline setting. As such, and as Judges Posner and Easterbrook discuss in their dissent, an improper application of the modified categorical approach would have implicated no Sixth Amendment concerns. Still, because the determination of a "violent felony" for ACCA purposes and a "crime of violence" for career offender purposes universally relies on the same (Taylor, Shepard, James, and Begay) analysis, it should stick in ACCA cases.

Posted by: afm | Mar 3, 2010 2:14:43 PM

afm, I think you are generally correct on your description and purpose of the modified categorical approach. However, one sentence caught my attention as being subject to mis-interpretation: "The modified categorical approach allows a sentencing court to go behind the actual words of a statute."

The MCA allows a court to only look and determine which part of the statute a defendant was convicted under. This is not an inquiry into the defendant's conduct, but rather an examination of the statute, as it was charged. I think we are both in agreement on this, but the passage from you I quoted above gave me some pause.

I think you'll find that other Circuits have taken this approach, in addition to the 7th in Woods. Check out US v. Zuniga-Soto, 527 F.3d 1110 (10th Cir. 2008), which is from the Circuit I'm most familiar with.

Posted by: DEJ | Mar 3, 2010 3:15:58 PM


I agree with you completely. What I meant by my statement was simply that a sentencing court may look at the Shepherd documents (in addition to the statutory language) to determine what part of the statute the defendant violated. Although my comment was not precisely worded, I did not mean that a court may consider the conduct those documents reveal.

Taylor provides an excellent hypothetical to illustrate my point. The Massachusetts statute at issue there was burglary. It included, in addition to burglary of a dwelling, burglary of a boat, and burglary of some other structure I can't immediately recall. Since the conviction was simply "Burglary," the court was permitted to look at other documents to rule out the possibility that the defendant had been convicted of burglary of the boat or the other specified structure.

So the MCA (thanks for the abbreviation) allows a court to look at documents other than the statute--which often reveal facts about the conviction--to determine what crime the defendant committed. The facts cannot play a role in the court's sentencing consideration, but they can be used to aid the court in determining whether the defendant has a prior predicate conviction.

Posted by: afm | Mar 3, 2010 3:37:42 PM

The foregoing discussion is both accurate and complete, especially afm's commentary, and I'm not sure I can add much to it, except like any attorney I'm bound to try. I think Johnson will help by eliminating many of the inconsistencies between and within (!) circuits. With proper use by defense counsel and application by the courts, we may see fewer of the especially glaring violations by sentencing courts of the plain meaning of the line of cases from Shepard. Judicial factfinding, which is patently impermissible, was all too common up to now, and we can hope that it will be diminished by Johnson.

Taken with Chambers, it is clear now that the statute of conviction has to define a violent crime, in whole (categorical approach), or in part (MCA). While Woods from the 7th Circuit does a pretty good job clarifying this analysis, unfortunately Johnson falls short here. The cite to Simms from the 4th Circuit in dicta is troubling, because the statute involved is one which pretty clearly doesn't qualify for the MCA, to me anyway, but I'm a defense attorney in the 4th.

Then again, the 4th handed down a decision in Rivers last week (Feb 25) which does just what the 7th did in Woods, though not as completely, by carefully analyzing the statute of conviction and determining that it does not define, in any of its constituent sections, a crime which can be conmmitted violently, and only violently, and therefore cannot be analyzed through the MCA. Now, this type of "cannot" determination, we hope, will be bolstered by a cite to Johnson and Chambers.

What also is troublesome here is that the MCA still remains a concept with ill-defined margins. The SCT and the circuits, when discussing the proper use of the MCA, have yet to lay down adequate boundaries regarding what can be considered in this and when, depending on the mechanism of the prior conviction -- plea, jury trial, bench trial. Sure, definitions have been set out, in Shepard for example, but when the "use of sanctioned documents" has been referenced to in subsequent opinions, the lines are blurred by imprecise recapitulation of the prior language, which I submit is occasioned by an inadequate understanding of the principle on the part of the circuits. For example, read carefully there can be no question that, per Shepard, charging documents only can be used when there is a plea, since after a trial they do not represent what are in any way agreed-upon facts. A trial is the ultimate contesting of facts, isn't it? But you can find many post-Shepard cases which do not make this clear, and which suggest that charging docs can be consulted after a trial. There will be more litigation, and more defendants will be taking appeals and habeas petitions, because of this imprecision.

Johnson will help, no doubt, but it will be as much due to good lawyering and a willingness on the part of the courts to honor its spirit and not just seek to distinguish its language. I don't think we can take any of that as a given. I will be interested to read some of the circuit cites noted in prior posts to perhaps help me in crafting better arguments. Thanks to everyone.

Posted by: mdm | Mar 4, 2010 9:39:58 AM

It seems like what I'm hearing is that this is to some degree just a restatement of the limitations of the modified categorical approach, but nevertheless a necessary and welcome one, as that doctrine tends to leak under pressure from aggressive ADAs in the lower courts, absent reinforcement from above.

That sounds good to me.

Posted by: Observer | Mar 4, 2010 12:07:07 PM

Not just the ADAs. Defense attorneys try the same thing too when their client's actions aren't that violent in the broad scheme of things, even if the typical means of committing an offense is violent.

In the Sixth Circuit, a defense attorney succeeded in doing just that--but thankfully the proper test was clarified in U.S. v. Young.

Posted by: Res ipsa | Mar 4, 2010 2:13:28 PM

So...is there any hope of judicial relief for inmates who raised the issue of being a violent offender" or are they out of luck?

Posted by: Renee | Apr 28, 2010 6:06:54 PM

Post a comment

In the body of your email, please indicate if you are a professor, student, prosecutor, defense attorney, etc. so I can gain a sense of who is reading my blog. Thank you, DAB