Qualifying Felonies

Qualifying Felonies

The felony-murder substitution described in the preceding sections does not attach to every crime the legislature has labeled a felony. A frequent mistake is to read the rule as reaching any felony that happens to precede a death, and that reading fails because of what the substitution depends on. Felony intent can stand in for malice only because willingness to commit a particular crime reveals indifference to human life, or at least gives the offender fair warning that lives are at stake. Where the felony carries little risk of death, that link is missing, and applying the rule would make a murder conviction turn on the accident that a death followed an ordinary offense. The felon's own choice would then explain none of the difference between a theft charge and a murder charge.

The decision that most clearly draws this line is People v. Phillips. The prosecution's theory was grand theft by false pretenses: the defendant had induced the parents of eight-year-old Linda Epping to cancel her scheduled cancer operation and place her under his care, and she died.¹ The trial court instructed the jury that a death occurring during a course of conduct amounting to grand theft was second-degree murder even though the death was not intended.² The California Supreme Court held that instruction erroneous, because only a felony inherently dangerous to life can support felony murder and the prosecution itself admitted that grand theft is not such a felony.³ The court added that the doctrine is a highly artificial concept that deserves no extension beyond its required application, and it noted that England, where the rule began, has abandoned it and that it has drawn severe and sweeping criticism.⁴

That holding states the main modern restraint. Only a felony inherently dangerous to human life can support the rule, and the danger is measured from the elements of the felony in the abstract, not from the particular facts of the case.⁵ The prosecution in Phillips asked the court to look at the defendant's entire course of conduct and count whatever dangerous features it contained. The court declined, because doing so would discard the statutory definition of the felony and replace it with the facts of the defendant's own behavior.⁶ A fact-based test would also collapse into asking whether this death was a foreseeable result of this conduct, which is close to the question ordinary implied-malice murder already puts to the jury, and the felony-murder shortcut would no longer shorten anything.

The dangerous-felony limit is not the only way statutes sort the felonies that count. Phillips observed that some felonies outside those enumerated in the California Penal Code's murder statute can still support a murder conviction under the felony-murder rule, which is the second-degree form the instruction in that case invoked.⁷ Statutes therefore add a layer on top of the common-law limit by naming particular felonies and attaching a higher degree of murder to them. The next three sections take these ideas in order: first the requirement that the felony be inherently dangerous to human life, then the abstract-elements test that measures that danger, and finally the enumerated felonies and the degrees of murder that statutes build around them. The first of these begins with the requirement that the felony be inherently dangerous to human life.

Inherently Dangerous Felonies

At early common law, a person whose conduct caused an unintended death during a felony or an attempt to commit one was guilty of murder, without regard to the dangerous nature of the felony or the likelihood that death would result from the way it was carried out.⁸ Older American formulations described the same idea by the character of the felony. The intent behind the killing had to be equivalent in legal character to a criminal purpose aimed against life, and that generally meant an intent to commit a specific felony, or at least to do "an act involving all the wickedness of a felony."⁹ Such a formula asked how wicked a crime was in general and never asked how likely it was to kill, which left the rule with no built-in limit once the list of felonies began to grow.

That breadth was tolerable only because of the setting in which the rule arose. Few felonies existed at common law, all were serious wrongs, most endangered life, and all were punishable by death, so it made little practical difference whether the felon was executed for the felony or for the death it produced.¹⁰ Legislatures later created many comparatively minor felonies, and the old formula then imputed malice from conduct far below the gravity of murder.¹¹ Take a statute that makes the sale of intoxicating liquor a felony. A buyer who drinks too much, falls asleep in a blizzard on the way home, and dies of exposure would not have died but for the sale, yet no one would call the seller a murderer. A taxpayer who files a false return is in the same position if the revenue agent investigating it slips on the front steps and breaks a neck.¹²

Courts and legislatures responded by trimming the rule, and the most widespread trim asks whether the felony is itself dangerous to life.¹³ The idea has an English pedigree. In the 1887 case that became the basis of the modern English rule before Parliament abolished it, Judge Stephen told the jury that it would be reasonable to confine the doctrine to acts known to be dangerous to life and likely in themselves to cause death, done for the purpose of committing a felony.¹⁴

The danger requirement has two supports, and both fail together when the felony is not dangerous. The first concerns what the felony intent reveals. Treating the intent to commit a felony as malice makes sense only if that intent shows a conscious disregard for human life, so where the acts that make up the felony carry too little danger to life, forming the intent to commit it establishes no such state of mind and the doctrine has no footing.¹⁵ The second concerns notice. When society declares dangerous conduct felonious, an offender cannot excuse a death by claiming ignorance of the danger, because the declaration itself warned of it.¹⁶ A felony without inherent danger gives no such warning, so the felon will not expect injury or death to flow from it, is unlikely to be deterred by the prospect of murder liability, and cannot logically be charged with malice.¹⁷ A structural reason reinforces both. Because proof of a qualifying felony and a connected killing settles malice as a matter of law, every felony added to the list places another set of killings outside the jury's judgment about the defendant's attitude toward life. That is why California gives the doctrine the narrowest application consistent with its purpose of deterring felons from killing negligently or accidentally.¹⁸

On the qualifying side sit the felonies that standard statements of the rule list, among them rape, arson, robbery, and burglary.¹⁹ Robbery is the clearest case, since a killing in the perpetration of robbery is first-degree murder even when it is inadvertent or accidental, which is why the death from fright in People v. Stamp could be charged as murder without any inquiry into the robbers' attitude toward the victim's life.²⁰ Burglary qualifies for a related reason, because California's burglary statute is aimed mainly at forestalling dangerous confrontations between intruders and occupants.²¹

The cancer-cure fraud in People v. Phillips marks the other pole. The prosecution conceded that grand theft by false pretenses is not inherently dangerous to life, and the California Supreme Court refused to extend the felony-murder rule to it, observing that it knew of no case in any jurisdiction applying the doctrine because death resulted from a course of conduct involving a felonious fraud.²² The deadly consequences did not change the analysis, because the inquiry concerns the felony's own character. The firearm cases draw the same line. In State v. Anderson the Minnesota Supreme Court held that a felon's possession of a firearm, and possession of a stolen firearm, are not inherently dangerous, because simple possession standing alone involves no special danger to human life even though use of a firearm can.²³ Robbery, by contrast, involves a taking by force or fear, so an element of the offense itself carries the threat of violence.²⁴

No fixed list settles which felonies qualify. Courts decide felony by felony, and they sometimes disagree. California decisions have held shooting at an inhabited dwelling, poisoning with intent to injure, arson of a motor vehicle, manufacturing methamphetamine, and kidnapping to be inherently dangerous, while holding grand theft, extortion, escape, and possession of a sawed-off shotgun not to be.²⁵ Drug sales divide courts as well, since a few jurisdictions treat them as inherently dangerous and most hold that sales are not dangerous in themselves.²⁶ Even the wording of one statute can decide the question. In People v. Howard the California Supreme Court held that driving with wanton disregard for safety while fleeing a pursuing officer is not an inherently dangerous felony, because the single statute also reaches traffic-point violations that put no one at risk.²⁷ Justice Baxter, dissenting, reasoned that such reckless flight is dangerous by definition and gives clear warning that a death may bring murder liability.²⁸ Courts also differ on how danger is measured, since Minnesota weighs the circumstances of the particular felony along with its elements, while California looks to the elements alone, the method taken up in the next section.²⁹

Failing the dangerous-felony test closes only the felony-murder route. Students lose this issue when they conclude that a non-qualifying felony means no murder liability at all, because the prosecution may still try to prove malice through the ordinary routes if the facts support it. The instruction in Phillips shows the structure. It had three parts: an unlawful act dangerous to life performed by a person who knows it endangers life, an abandoned and malignant heart, and a killing in the perpetration of a felony such as grand theft.³⁰ The first two are ordinary implied-malice theories, and only the third was error. That error was prejudicial because it let the jury convict on proof that the death proximately resulted from grand theft, without finding express malice or the implied malice grounded in an intent, with conscious disregard for life, to commit acts likely to kill.³¹ The prosecution's remaining path was to prove that mental state to the jury directly.

The same pattern appears where the facts are worse for the defendant. In People v. Sanchez the court held that a violation of an evading-police statute, section 2800.3, is not inherently dangerous, because its dispositive elements can be met without a high probability of death. It also found overwhelming evidence of implied malice in an intoxicated driver with a prior DUI conviction who raced through residential streets at up to 100 miles per hour.³² The court still treated the felony-murder alternative as a real problem, because the prosecutor had told the jurors the theory could shorten their deliberations and nothing in the general verdict ruled out that they took the shortcut.³³ Likewise, the high-speed chase and fatal collision in Howard could readily have supported a conviction on ordinary depraved-heart malice.³⁴ The felony-murder rule is a shortcut around proving malice, so a predicate felony that fails the test removes the shortcut and leaves the definition of murder untouched.

The Abstract Elements Test

Requiring that a predicate felony be inherently dangerous to human life raises an immediate follow-up question: dangerous according to what? One measure looks at the offense as the legislature defined it and asks what the elements demand of anyone who commits it. The other looks at the offense as this defendant committed it and asks whether the particular conduct put lives at risk. California adopts the first measure. A court deciding whether a felony can support felony murder examines the elements of the felony in the abstract, meaning the statutory definition taken apart from the facts of the case, and it does not consult the defendant's specific conduct.³⁶ The reasons for that choice show most clearly in the case where a prosecutor tried to use the other measure.

In People v. Phillips the prosecution admitted that grand theft is not inherently dangerous to life. It then asked the court to look past that admission by taking in the defendant's entire course of conduct, so that whatever dangerous features the conduct contained could be counted. The Attorney General proposed to describe the offense as "grand theft medical fraud," a newly created felony that plainly endangers life.³⁷ The California Supreme Court refused, because defining the felony by the defendant's actual behavior would abandon the statutory definition and put the facts of the case in its place. The court also saw that if any dangerous segment of a course of conduct could count, the rule would reach every felony during which the defendant happened to act in a way that endangered life. Once the legislature's definition was set aside, nothing would limit the contextual features that could be folded into an expanded definition, so the court declined to enter that territory.³⁸

The elements-only approach rests on two supports, and the structure of the rule supplies both. The first is a limit on the prosecution's reach. Almost any felony can be committed in some risky way, and a label describing that risk can always be invented after a death has occurred, so a test that accepts facts and labels hands the prosecution control over which felonies count. The legislature's definition is the one boundary the prosecution cannot redraw after the fact. The second support is notice. The theory behind imputing malice is that when society declares dangerous conduct felonious, an offender has been warned of the risk and cannot claim ignorance of it.³⁹ That warning can come only from the text of a statute, which exists before the crime, while the facts of how the defendant behaved do not exist until the crime is over. A person deciding whether to commit an offense can learn whether it carries murder exposure only if the test turns on what the statute says.

The third step asks what the statute permits, which is a different question from what offenders usually do. In People v. Sanchez the People pointed out that a violation of the statute on fleeing a pursuing officer and proximately causing death or serious injury usually involves a high probability of death. The court found that no help, because a felony qualifies only if its dispositive elements, viewed in the abstract, cannot be satisfied without creating a substantial risk that someone will be killed. The elements of that statute can be met by conduct that does not necessarily carry a high probability of death.⁴⁰ Frequency is a fact about how offenders tend to behave, and the test consults only what the definition allows. People v. Howard shows the same discipline at the level of a whole statute. The California Supreme Court treated the offense of driving with wanton disregard for safety while fleeing an officer as one felony that also covers traffic-point violations, and it held the felony not inherently dangerous because one way of committing it puts no one at risk.⁴¹ Justice Baxter, dissenting, would have analyzed only the reckless-driving subdivision, since the prosecution charged and the jury was instructed only under that language.⁴² The majority's approach keeps the unit of analysis at the offense as the legislature wrote it, so a charging choice cannot carve a dangerous version out of a broader statute.

The abstract test can also qualify a felony, and the example here is an illustration rather than a decided case. Take a hypothetical statute that makes it a felony to set fire to a dwelling while people are inside. Occupancy is part of the definition, so every violation puts occupants at risk, and a court reading only the elements would call the felony inherently dangerous. It would not matter that one arsonist set the fire at noon beside a firehouse, because the facts of the particular commission are never consulted. California decisions reach the same result on actual offenses such as shooting at an inhabited dwelling and manufacturing methamphetamine.⁴³ Set that against Phillips, where the elements of grand theft carry no danger to life and the defendant's dangerous course of conduct could not repair the gap. The same sort of dangerous facts produced opposite results because the elements, and not the facts, did the sorting.

Not every jurisdiction uses that measure, so the first task on an exam is to identify which one the problem follows. Minnesota weighs the elements of the underlying felony in the abstract together with the facts of the particular case and the circumstances in which the felony was committed.⁴⁴ Under that approach a property crime can qualify. In State v. Cole an armed defendant tried to exchange stolen goods for cash at a department store and shot a police officer to avoid arrest, and the court held that both the theft and the assault were proper predicates because the circumstances showed a special danger to life.⁴⁵ A theft felony that could not support felony murder in Phillips thus supported it in Cole, and the only difference was the measuring stick. Minnesota still looks at the abstract side too, which is why it held that a felon's possession of a firearm is not inherently dangerous, since simple possession carries no special danger by itself.⁴⁶ Some states also ask whether the act that caused the death was dangerous, which is a different inquiry from whether the felony is dangerous. A fraud that induces a victim to forgo medical care may be an extremely dangerous means of committing a felony that is not dangerous in itself.⁴⁷

Where a problem applies the California approach, the common error is to read the stem for risky behavior and let it decide the question. The fact pattern may stress how dangerously the defendant actually behaved, but the test discards that narrative and asks only what the statute requires of every offender. If the offense fails on its elements, the defendant is not automatically freed from murder liability, because the prosecution may still try to prove malice through the ordinary routes.

Enumerated Felonies and Degrees

The inherent-danger requirement and the abstract elements test decide whether a felony can supply malice at all. A further question follows once it can: what grade of murder results? Many jurisdictions answer that question by statute, and California's Penal Code section 189 shows the pattern. It makes murder first degree when it is committed by specified means such as poison or lying in wait, when it is a willful, deliberate, and premeditated killing, or when it is committed in the perpetration of, or attempt to perpetrate, one of the felonies the section lists, and it makes every other murder second degree.⁴⁹ A felon who kills therefore lands in one of two places. If the felony is on the statute's list the murder is first degree, and if the felony is inherently dangerous but unlisted the murder is second degree.⁵⁰

The two-level structure has a history that explains its shape. California's first criminal statute, enacted in 1850, defined murder as an unlawful killing with malice aforethought, recognized a single degree of murder punishable by death, and carried a proviso that kept killings during a felony out of involuntary manslaughter, which in effect codified the common-law felony-murder rule.⁵¹ In 1856 the Legislature divided murder into two degrees and assigned murder committed in the perpetration of listed felonies to the first, leaving all other murders in the second.⁵² For a killing by a listed means such as poison, section 189 only fixes the degree, so malice must first be proved independently, while for a homicide resulting from a listed felony California decisions treat the section as also codifying the felony-murder rule, so that no independent proof of malice is needed and the killing is first degree as a matter of law.⁵³ The California Supreme Court has summarized the resulting scheme as one in which the rule posits malice for homicides that directly result from all inherently dangerous felonies, and posits malice and also classifies the offense as first degree for the felonies enumerated in section 189.⁵⁴

Robbery is the clearest example of the first-degree route. Statutes that grade murder by degree have in most cases kept premeditated murder and some form of felony murder as alternative ways to reach the first degree, and robbery is invariably among the felonies covered.⁵⁵ In People v. Washington the California Supreme Court restated what that means: although section 189 speaks only of degrees of murder, inadvertent or accidental killings are first degree murder when felons commit them in the perpetration of robbery.⁵⁶ That case also held that a killing committed by the robbery's victim falls outside the rule, a limit taken up in a later section.⁵⁷ For a killing by the felon or an accomplice, liability requires an intent to commit the underlying felony, and it requires neither an intent to kill nor implied malice.⁵⁸ Premeditation drops out as well, since first-degree murder otherwise requires an intent to kill formed after premeditation and deliberation.⁵⁹

Each part of this scheme has its own reason. Substituting the felony for premeditation is a grading choice: first-degree murder ordinarily marks a killer who planned the death, and a legislature that wants the same grade for killings in the most dangerous felonies has no planning to point to, so the statute points to the felony instead. In People v. Burton the California Supreme Court described the choice as the Legislature's judgment that the deterrent purpose of the rule outweighs the usual policy of examining each killer's state of mind and calibrating the punishment to it.⁶⁰ The deterrent target is the negligent or accidental killing, which the rule tries to prevent by holding felons strictly responsible for the killings they commit.⁶¹ The choice of which felonies go on the list rests on a second reason. Robbery and the other listed felonies are pursued for a purpose independent of the homicide, such as acquiring another's property, and a person who embarks on one of them comes within a clear legislative warning that a resulting death will be first degree murder regardless of the circumstances.⁶² Because the statute fixes the list before any crime occurs, an offender can learn the consequence in advance, and advance knowledge is what gives a deterrent rule its chance to work.

Felonies outside the list reach felony murder by a different and lower route. The first-degree rule is a creation of statute, while the second-degree rule is a common-law doctrine covering an unlawful killing in the course of an inherently dangerous felony that section 189 does not enumerate.⁶³ The second-degree rule is itself tied to the statutory definition of implied malice, because the willingness to commit a felony inherently dangerous to life is a circumstance showing an abandoned and malignant heart.⁶⁴ That tie explains the lower grade. The common law supplies malice for the unlisted felony, which is enough for murder, but nothing in the statute raises the grade, and section 189 leaves every murder it does not name in the second degree.⁶⁵ The inherent-danger inquiry from the earlier sections therefore does its main work for unlisted felonies, because for a listed felony the Legislature has already made the judgment and has decided the felon is no longer entitled to individual calibration of the offense.⁶⁶

The effect of the list shows most clearly where a similar wrong can fall on either side of it. In People v. Phillips the prosecution proceeded on grand theft by false pretenses against a defendant who falsely claimed to cure cancer, and the court held that a felony-murder charge can rest only on a felony inherently dangerous to life and that grand theft is not such a crime.⁶⁷ Change the felony through which a similar theft is carried out and the analysis changes. In People v. Gonzales the California Supreme Court noted that a defendant who enters a store intending to commit theft by false pretenses, and who kills an employee while fleeing, can be convicted of felony-murder burglary.⁶⁸ Burglary is among the felonies section 189 lists, and it requires only an entry with the required intent, so the entry turns a plan for theft by deception into a listed felony even though nothing is taken.⁶⁹ The fact that moved the answer was whether the felony charged appears on the list.

None of this is uniform from state to state. Statutes that enumerate felonies for first-degree felony murder commonly name arson, burglary, kidnapping, rape, and robbery and treat every other felony murder as second degree, but each legislature writes its own list.⁷⁰ In the District of Columbia, unintentional killings are first degree only when they occur in one of six enumerated felonies, and the status of an unintentional killing in an unlisted felony remains unclear.⁷¹ Even within a state's own list, the defendant must have intended to commit the listed felony when the killing occurred, so an attack unaccompanied by an intent to commit mayhem does not become first-degree murder merely because mayhem is listed.⁷²

Florida shows that the degree can also turn on the defendant's participation. Its first-degree statute covered a killing committed by a person engaged in the perpetration of, or attempt to perpetrate, robbery and other listed felonies, and the trial judge in Enmund v. Florida told the jury that such a killing was first degree even without a premeditated design or an intent to kill.⁷³ The judge then drew the degree line by participation. First-degree felony murder required proof that the defendant was actually present and actively aiding and abetting the robbery, while second-degree felony murder required proof of a killing in the perpetration of the robbery and that the defendant, though not physically present, procured, counseled, commanded, or aided another to commit the crime.⁷⁴ A codefendant tried separately, Jeanette Armstrong, was convicted of two counts of second-degree murder.⁷⁵ Enmund was convicted of first-degree murder, and although the Florida Supreme Court found no direct evidence that he was at the back door when the killings occurred, it rejected his argument that at most he was guilty of second-degree murder, reasoning that felony murder and the law of principals combine to make a felon generally responsible for a co-felon's lethal acts.⁷⁶ The death sentence imposed on Enmund raises a separate constitutional question, taken up in its own section below. For present purposes the case shows that a statute can grade by participation, so that the same robbery and the same death may support first-degree liability for one participant and second-degree liability for another, and that rule is Florida's rather than a general one.

Students most often carry one state's list into another state's problem, or assume that an accident lowers the grade. For a listed felony the accident changes nothing, because the statute grades the killing by the felony, whatever the felon planned. The list and its degree conditions belong to the governing statute, so an exam stem that supplies statutory language controls over any list remembered from California or Florida.