Concept

Analyzing a Causation Problem

Proximate cause and divergence of result

Analyzing a Causation Problem

The preceding sections taught the Code's causation rules one at a time, but an exam fact pattern does not announce which rule applies, so a problem has to be worked in a fixed order. The order comes from the structure of § 2.03(1): conduct is the cause of a result only if it is a but-for antecedent of that result and, in addition, the relationship between conduct and result satisfies any further causal requirements imposed by the Code or by the law defining the offense.¹ The first layer is a factual question with a yes-or-no answer. The second layer is where the divergence rules of subsections (2), (3), and (4) operate, and it asks how far the actor's culpability reaches, because the Code converts much of what the common law called proximate cause into an inquiry about culpability toward the result.² That conversion explains why the second layer cannot be answered until the offense's mental state has been identified. The call of the question also deserves a first look, since it may state whether the jurisdiction follows the Model Penal Code or the common law, and the sequence below assumes the Code.

A fixed order also guards against two opposite mistakes. The first treats a harmful result as attributable to the defendant once the conduct is shown to be a but-for antecedent, which overlooks how far that test reaches: almost every harm has some remote antecedent in someone's conduct, so but-for alone cannot separate fair attribution from absurd attribution.³ The second treats a freakish chain of events as automatically excusing the defendant, which overlooks that the Code excuses only results falling outside the baseline the mental state fixes and forgives a good deal of variation inside it.⁴ Taking the questions in order puts each one to the facts at its turn. The prosecution carries the burden throughout, because causing a result is an element that must be proved beyond a reasonable doubt whenever the offense requires it.⁵

Exam Analysis Steps

Because each later question depends on the answer to an earlier one, the analysis is best run as a sequence in which every step either ends the problem or hands it to the next. The steps are these:

Confirm that the defendant's conduct is a but-for cause of the result by asking whether the result would have occurred had the conduct not taken place.
Identify the result element the offense requires and the mental state attached to that element, because that mental state (purpose or knowledge, recklessness or negligence, or none) selects which subsection of § 2.03 supplies the test.
Compare the actual result with the baseline that mental state supplies, which is the designed or contemplated result for purposeful and knowing offenses and the probable result of the risk the actor was aware of, or should have been aware of, for reckless and negligent ones, and stop if the actual result falls within that baseline.
If the results differ, ask whether the only difference is that a different person or different property was injured or affected, or that the harm was less serious or less extensive than the baseline, and treat the element as established if it is.
If the difference is anything more, ask whether the actual harm is the same kind of injury or harm as the baseline and is not too remote or accidental in its occurrence to have a just bearing on the actor's liability or the gravity of the offense, and treat the element as established only if both conditions are met.
If the offense imposes absolute liability for the result, skip the comparison and ask whether the actual result was a probable consequence of the actor's conduct.

Most problems end at the third step. In the usual case the defendant aims at a healthy victim and hits that victim, or drives recklessly and kills another motorist in a crash, and the actual result is exactly what the baseline describes, so nothing further needs to be said.⁶ The unusual patterns that signal work at the fourth and fifth steps are familiar ones: a shot that misses its target and hits a bystander, a victim weakened by sickness, negligent medical treatment of a wound, a victim who commits suicide to escape the pain, and a victim who dodges a bullet and drowns in a river.⁷ The wrong-victim pattern is sorted at step four, and the others are tested at step five. For a purpose-based crime the same sequence takes a compact form: the actual result, whether it was within the actor's purpose, whether it involved the same kind of harm, and whether it was too remote or accidental. For recklessness and negligence, the probable result of the risk simply replaces the designed result as the point of comparison.⁸

One fact pattern shows how the steps hand off to one another.

Step one is satisfied because the operator would not have been on the highway had the thief not stolen and fled. At step two the result element is the death of another person and the attached mental state is recklessness, which routes the problem to subsection (3) and makes the baseline the risk the thief was actually aware of, here that running across a fast road in the dark would get someone on it hit by a vehicle. At step three that baseline, a person struck by a car and gravely hurt or killed, is compared with the actual result, a person struck by a car and killed, and the two nearly coincide. The one gap is that the victim is someone other than the people the thief had in mind, such as the thief or the motorists, and step four resolves it because a change in the identity of the person harmed is the first variation the recklessness rule forgives.⁹ Because the analysis ends there, the operator's decision to give chase and the driver's conduct never have to be examined as remote events. A careful answer still adds a fallback at step five. An intervening act is treated as dependent on the defendant's conduct when it was foreseeable or sufficiently related to it, and pursuit of a fleeing thief is the kind of reaction a theft invites.¹⁰ A passing driver is likewise among the ordinary hazards of a highway. Courts reason the same way about roadside hazards: in People v. Ryan, an appellate court upheld a finding that a drunk driver's collisions set in motion the events that killed a police officer struck by another driver at the scene, because it was reasonably foreseeable that the collisions would endanger police and other first responders.¹¹

Hypothetical Illustration

In a variation on the same theft and flight, the operator crosses the highway safely behind the thief and is killed on the far shoulder when lightning strikes a nearby tree. The theft and flight are still a but-for cause, since the operator would not have been standing there otherwise, and death is still bodily harm of the kind the highway risk presented. Step four cannot be met, however, because the actual result differs from the baseline in more than the identity of the victim: it arrives by a route the risk did not include. Step five then asks whether the death is too remote or accidental to bear on the thief's liability, and a fact-finder may find that it is, because the lightning is a coincidence the flight did nothing to make likelier. If so, the element of recklessly causing death is not established.

The two scenes share a defendant, a baseline, and a victim, and they differ in a single fact, the route to the harm. That fact moves the problem from step four to step five, because the different-victim clause applies only when the identity of the victim or property, or a milder extent of harm, is the sole divergence.¹² Step one gave the same answer in both scenes, which shows how little work but-for causation does by itself. The real decisions came at steps three through five.

Step six replaces steps three through five instead of following them. An offense of absolute liability attaches no mental state to the result, so there is no designed or risked result to compare, and the Code supplies a different additional causal requirement of the kind the second layer of § 2.03(1) contemplates: the result must be a probable consequence of the actor's conduct.¹³ The license example arises in the setting of misdemeanor manslaughter, which is taught elsewhere.

In the license example step one is met and step six decides the problem. The contrast with the thief's flight shows what step six measures: the flight made a fatal accident on a busy road a real prospect, while lapsed paperwork makes none, so the death falls outside what the conduct could be expected to produce.

If the jurisdiction applies common-law proximate cause instead of the Code, the same facts are sorted by three questions: whether the result was an objectively foreseeable consequence of the conduct, whether the conduct is an efficient cause, meaning the chain of events is short enough to warrant punishment, and whether an independent, free-willed agent intervened and broke the chain.¹⁴ Foreseeability and the length of the chain correspond to the remote-or-accidental condition at step five, and the free-willed-agent question corresponds to the intervening-act analysis there. The Code reaches the same concerns through the actor's culpability, which is why its sequence runs through the mental state.¹⁵

The most common error is to jump from the but-for finding to a comparison of results without first fixing the mental state attached to the result. The baseline comes from that mental state, so a comparison made against the wrong baseline produces the wrong answer. A purposeful-killing charge measures the result against what the actor designed, a reckless one measures it against the risk the actor was aware of, and an absolute-liability offense has no mental baseline at all.¹⁶

Analyzing a Causation Problem

Endnotes17