William Lane Craig’s Legal Analogies: Can They Justify Penal Substitution?
In much of Western Christian thought, particularly in Protestant circles, penal substitution is the most prominent theory of understanding the atonement. According to this model, Jesus was punished in the place of sinners. He bore the penalty they deserved, satisfying the demands of divine justice so that God could righteously forgive those who believe in Him.
But this theory faces a primary objection: How can it be just to punish an innocent person?[1]
In response, Christian philosopher William Lane Craig has turned to the concept of legal fiction to defend the coherence and justice of penal substitution. A legal fiction is a practice in human law where something is treated as true even though it is not factually so for the sake of legal functionality. For example, a corporation is treated as a “person” in many legal systems, even though it is not human. Courts may also treat an adopted child as a biological heir for inheritance purposes. These fictions allow the law to operate consistently and effectively, even if they do not reflect the actual facts of the situation.
Craig argues that something similar occurs in the atonement. Jesus is not actually guilty, but God imputes the guilt of humanity to him. God treats him as if he were guilty and punishes him accordingly. In the same way, Jesus’s righteousness is imputed to believers, even though they are not actually righteous. Craig claims that this transfer of legal status is morally justifiable and appeals to the widespread and accepted use of legal fiction in human jurisprudence to make its acceptance more defensible. Craig acknowledges that legal fictions are not factually accurate but argues that they are still legally valid. He says that since God is the ultimate moral authority, He has the right to impute sin and treat Jesus as guilty for our sake.
He also emphasizes that Jesus willingly accepts this role, which supposedly preserves the moral integrity of the act. Craig also appeals to the legal doctrine of vicarious liability, especially in tort law, where a party (like an employer) can be held legally responsible for the actions of another (like an employee), even if the party itself did not commit the wrongful act. Craig uses this to argue that moral or legal liability can, in principle, be transferred or imputed from one party to another, justifying how Jesus could bear our sins and their punishment. He argues this notion of representation allows Jesus to bear our penalty in a morally coherent way. Craig would stress that he is not a disconnected bystander but a God-incarnate, assuming human responsibility and fulfilling divine justice through mercy.
However, while legal fiction works to solve procedural problems in human systems, God is not a bureaucrat. Human courts use fiction because they are limited by complexity, uncertainty, and the need for consistent procedures. God, by contrast, is morally perfect, omniscient, and not bound by such limitations. If divine justice is truly just, it must correspond to reality, not simply declare things into being for convenience. Transferring sin or righteousness legally doesn’t mean there’s real guilt to punish. Thus, punishment remains unjustified unless the guilt is genuine. Penalizing an innocent, even by divine fiat, still violates the very concept of just punishment, which necessarily expresses the deserved infliction of suffering/condemnation of a wrongdoer. Contingent legal definitions don’t resolve the incoherence. Craig’s legal fiction defense doesn’t deal with reality: moral guilt and legal guilt are not isomorphic. Without real guilt, punishment, even fiction, is hypocritical. Human judicial fictions serve practical ends; divine justice is absolute; it must correspond to reality, not mere pragmatism. Legal fictions (like treating a corporation as a “person”) are tolerated because they serve administrative efficiency, not because they’re morally sound analogies to the punishment of an innocent person.
Also, when it comes to vicarious liability in law, this is mostly (not only) civil, not criminal. A company might be held liable for an employee’s actions, but it isn’t sent to jail; it pays damages. In criminal law, punishing an innocent party for the crimes of another is almost universally considered morally outrageous and a miscarriage of justice. Legal fictions (like treating a corporation as a “person”) are tolerated because they serve administrative efficiency, not because they’re morally sound analogies to the punishment of an innocent person. Craig overextends these analogies beyond their appropriate domains. While he admits these are analogies, he treats them as if they substantiate moral legitimacy, not just conceptual coherence. His model asks us to embrace a view of justice in which God punishes the innocent to save the guilty, a vision many find ethically troubling. These elements fail to neutralize the basic concern: punishing an innocent person, even voluntarily, seems inconsistent with the nature of a just and merciful God.
Moreover, voluntariness does not remove injustice. If an innocent person volunteers to be punished for someone else’s crime, we do not consider that morally acceptable. Justice demands that punishment be directed at the guilty, not simply at someone who agrees to stand in their place. Craig’s model treats punishment as if it were transferable, like property, but guilt is not a transferable substance. Even Jesus’s voluntary assumption of guilt does not align here because criminal law does not allow volunteers to be punished for others’ crimes. A court cannot let a volunteer go to prison in place of a rapist or murderer. This is morally distinct from attributing blame in corporate law, as those laws aim to enforce compliance and prevent harm, not to reflect desert or to morally condemn someone blameless.
With that said, we can better assess Craig’s analogies from Allen v. Whitehead (1930) 1 K.B. 211 and Sherras v. De Rutzen (1895) 1 Q.B. 918. Let’s walk through them and see if they truly support his claim that we have real-world analogs of guilt being justly imputed to a blameless person.
In the case of Allen v. Whitehead (1930) 1 K.B. 211, a café owner delegated management to someone who allowed prostitutes to congregate, which violated a statute. The court held the owner criminally liable, even though he did not directly commit or authorize the act. Craig uses this to show that criminal guilt can be imputed to someone blameless under the delegation principle, meaning Jesus bearing our guilt isn’t alien to human justice.
However, this is a case of regulatory or public welfare offenses where strict liability is applied. The liability is not moral guilt in the strong sense; it’s legal accountability meant to incentivize oversight. The law here holds the owner liable because he holds a position of responsibility and could have prevented the offense. This is not a case of someone being punished instead of the guilty person, but in addition to them, as a consequence of oversight duty.
So, this shows imputed legal liability; yes. But not the substitutionary punishment of the innocent in place of the guilty, which is essential to penal substitution.
In the case of Allen v. Whitehead (1930) 1 K.B. 211, a bartender sold alcohol to a police constable on duty. The license-holder (owner) was held liable because the law considered him the “true seller,” even if he was absent. Craig presents this under the attribution principle, where acts of one are treated as the acts of another, supporting the idea that Jesus could be treated as if he committed our sins.
Nevertheless, this is a strict liability in a licensing context. The law ‘fictionally’ attributes the bartender’s act to the licensee, not because the owner is morally responsible, but to ensure accountability in commerce. This is not punishment as a substitution, nor is it moral condemnation. It’s part of regulatory compliance, akin to a fine for breaking a license rule, not prison for murder. It shows legal fiction and functional attribution but again, does not mirror penal substitution, where one person is intentionally punished for another’s moral crimes substitutionally.
These cases support Craig only partially, and only in a narrow legal sense. They show that liability can be attributed to someone other than the agent. Blameless parties can be legally punished or fined, even in criminal law. While Craig is right that some legal traditions tolerate the imputation of liability to innocents, they fail to support the heart of Craig’s theological analogy, as they are not moral endorsements of such liability, not examples of substitutionary punishment, and not seen as ideal justice, even within the law; they’re sometimes criticized for violating moral intuitions.
If anything, these cases highlight the problem with penal substitution rather than solve it. The only human examples Craig can cite are strict liability offenses, which ignore mens rea, are applied in limited regulatory domains, and are often regarded as exceptions to justice, not exemplars of it.
If this is the best the analogy can offer, it arguably undermines the justice of penal substitution more than it supports it.
In law, the punishment is due to institutional responsibility, not moral substitution. In Christian theology, Jesus bears our punishment not because he’s institutionally responsible but a willing, personal substitute who absorbs divine wrath in our place. So, even though both scenarios involve the punishment of a non-actor, only penal substitution involves the claim that this is morally just and redemptive.
In Allen v. Whitehead, the owner isn’t being punished instead of the manager. The law says, “Because you own the café and delegated duties, you’re jointly liable for what happens under your watch.” However, in penal substitution, the claim is: “Because Jesus chose to take on your guilt, he is punished so that you don’t have to be.” These are fundamentally different moral structures.
Let’s provide a helpful analogy as to why this isn’t like penal substitution. Imagine a father whose child steals from a store. The store says, “Your child stole, so you must pay.” That’s like vicarious liability; the father pays the fine not because he stole, but because he’s responsible for the child. He’s not being morally blamed; he’s being held financially liable. Imagine the store says, “We’ll put your father in jail instead of you.” That’s like penal substitution. But that would be illegal in modern justice and not seen as just, even if the father volunteered.
That’s how one should view the difference between legal vicarious liability (what Craig cites), and theological penal substitution (what Christians believe happened on the cross). Craig’s legal cases do not involve true substitution; the wrongdoer is still liable. They involve functional liability, not moral guilt or deserved punishment. The punishment of Jesus is not shared with the guilty but taken on behalf of the guilty; that’s a unique and controversial moral claim. Therefore, legal vicarious liability cannot justify penal substitution; at most, it suggests that the idea of imputation exists but not that it is morally right or analogous in depth.
Moreover, in both Allen v. Whitehead and Sherras v. De Rutzen, the actual wrongdoer (the manager or bartender) did not get off the hook. The employer or license-holder was also held liable, but not instead of the wrongdoer.
In Allen v. Whitehead, a café owner delegated management to someone who allowed prostitutes to congregate. The manager was guilty because he knowingly allowed the illegal activity. The owner was also found guilty under strict vicarious liability because he delegated the legal responsibility for managing the premises. But the court did not absolve the manager. The owner’s guilt was additional, not substitutionary.
In Sherras v. De Rutzen, a bartender served alcohol to an on-duty police officer, which was illegal. The bartender was clearly the one who did the act. But, the law treated the license holder as the “true seller” for regulatory purposes. Again, both could be held liable: the bartender for his direct violation and the license holder for holding the license and failing to prevent misuse.
Craig’s defense of penal substitution relies on showing that our legal systems accept the just punishment of an innocent party in place of the guilty. But in these cases, there is no substitution, no moral exoneration of the actual wrongdoer, and the innocent party is not punished but alongside the actual wrongdoer due to a technical, legal responsibility.
The bottom line is that the employee who committed the crime remained liable and guilty in both cases. The employer or license-holder was not punished in place of them, but in addition to them, due to delegated or formal legal responsibility. This completely undercuts Craig’s analogy. It shows that these cases do not mirror penal substitution, where Jesus is punished so that we are not.
In conclusion, William Lane Craig’s use of legal analogies is an innovative attempt to answer a serious challenge to penal substitutionary atonement. But in the end, it creates more problems than it solves. It redefines guilt in a way that separates it from moral reality, justifies the punishment of the innocent through fallacious procedural analogies, and treats justice as a matter of legal status rather than moral truth.