The but-for test is a test commonly used in both tort law and criminal law to determine actual causation. The test asks, "but for the existence of X, would Y have occurred?"
In tort law, but-for causation is a prerequisite to liability in combination with proximate cause. In the absence of either of these, a party cannot be held liable.
Despite its ubiquitousness, the but-for test has a couple of problems:
- Tenuous Relations Between Actions: If someone looks hard enough, most things can be construed as a but-cause.
- For example, a party who is robbed while taking an alternative route home from work due to construction can credibly claim the mayor who was elected on promises to start this construction is the but-for cause of the robbery. For many, the connection here is too remote to justify liability.
- Over-Causation/Merged Causes: Sometimes there are too many but-for causes such that it is difficult to parse out any individual cause’s contribution to the whole.
- For example, if two people shoot another at the exact same time, it may be unclear which bullet causes the death.
Courts have taken a multitude of approaches to solve these issues. The field of torts solves the tenuous relationship issue by requiring proximate cause in addition to but-for causation for liability. Additionally, many jurisdictions also solve the merged causation issue in the field of torts through the use of joint and several liability which holds all contributing parties liable for the whole of the damage so long as they were negligent.
Criminal law has its own set of solutions. Some courts use the "substantial factor" test, which states that as long as a defendant's actions were a substantial factor in the crime, then that defendant can be found guilty. Under this test, both gunmen would be found guilty despite ambiguities in whose bullet caused the death. However, this test creates a problem in which the members of the firing squad whose bullets did not harm the victim are still guilty, even though their actions did not lead to the victim's death. Courts have developed four other tests to try and solve this issue.
- Likelihood of Survival Test: If the defendant's actions decreased the victim's chance of survival, then the defendant is guilty.
- Some courts have rejected this test because they believe that the intervening action only presents a mere possibility that the person’s life would have been saved. As such, we cannot be certain that the absence of that test actually contributed to the death (for example, under the likelihood of survival test, say that a doctor refuses to perform "x" surgery or test on the patient. After the doctor failed to perform the surgery or test, the patient died. Under the likelihood of survival test, the doctor would be found guilty, because performing the test would have increased the likelihood of the patient's survival. However, courts that reject the test would say that the doctor's performance of the test or operation would not necessarily increase the likelihood of survival, because the patient may have still had the same chance of dying, and so those courts would not find the doctor guilty.)
- Acceleration Theory: If the defendant's action caused a victim to die sooner than the victim would have otherwise died, then the defendant is guilty.
- For example, If "X" fatally poisons "Y," but "Z" shoots and kills "Y," under acceleration theory, Z is convicted, rather than "X."
- Proximate Cause: Some courts have scrapped but-for cause altogether, and simply apply the doctrine of proximate cause.
- Under this test, a defendant whose actions are closely enough related to the result is guilty.
- Model Penal Code (MPC) Approach: The defendant is guilty if the result of the defendant's action involves the same kind of injury or harm as the probable result, and the result is not too remote or accidental in its occurrence.
[Last updated in June of 2022 by the Wex Definitions Team]
- Keywords