Passage
A strict liability offence is one that may be committed without proof of any guilty state of mind in respect of at least one element of the offence. The defendant who sells food unfit for consumption, or who discharges a pollutant into a river, may be convicted whether or not she knew, or could reasonably have known, of the contamination. To many lawyers this is an affront. The criminal law, on the orthodox account, condemns the blameworthy, and blame presupposes a choice to do wrong or a careless indifference to the risk of doing so. An offence that dispenses with the guilty mind, the argument runs, is incompatible with the basic principles of criminal responsibility. The argument is powerful, but it proves less than it claims.
A. P. Simester has set out the orthodox objection in its strongest form. The criminal conviction, on this view, is not merely a mechanism for regulating conduct; it is a formal expression of social condemnation, and condemnation is intelligible only where the defendant could have acted otherwise. To convict a person who took all reasonable care, and who could not have avoided the prohibited outcome, is to condemn her for something that was, in the relevant sense, not her fault. The conviction says 'you are to blame' to a person who is not to blame, and a system that does this routinely has lost its grip on the connection between guilt and blameworthiness that gives the criminal law its moral authority.
The objection is at its most compelling when it is directed at strict liability for serious offences carrying real stigma and severe punishment. It is considerably weaker when directed at the regulatory offences where strict liability is in fact concentrated. Jeremy Horder has drawn the relevant distinction. Most strict liability offences are not crimes of stigma at all; they are regulatory provisions governing those who have chosen to enter a regulated activity (selling food, supplying water, running a factory) and who undertake, by entering it, to meet a standard the law sets in advance. The conviction of such a defendant carries little of the condemnatory freight that the orthodox objection assumes. It functions less as a judgment of moral character than as the enforcement of a standard the defendant accepted on entering the trade.
There is, moreover, a consequentialist case that the orthodox objection passes over too quickly. Strict liability, its defenders argue, induces a higher standard of care than a fault-based regime, because the operator who knows that reasonable care is no defence will take more than reasonable care. Findlay Stark has questioned how robust this effect really is, noting that the empirical evidence for the additional deterrent is thinner than its proponents assume, and that a due diligence defence (which permits acquittal on proof that all reasonable precautions were taken) captures most of the deterrent benefit while preserving the link to fault. The choice, on Stark's account, is not between strict liability and a regime that lets the careless escape; it is between strict liability and a carefully designed fault standard that does almost the same regulatory work without the same cost to principle.
What emerges is that the incompatibility thesis is true of some strict liability offences and false of others. Strict liability for a genuinely stigmatic crime, carrying imprisonment and the condemnation of the community, is indeed difficult to reconcile with the principle that criminal blame tracks fault. Strict liability for a regulatory contravention, carrying a fine and little stigma, imposed on a person who chose to enter the regulated activity, is a different matter, and the orthodox objection loses much of its force when transplanted to that setting. The error of the incompatibility thesis is not that it misunderstands the principle of criminal responsibility. It is that it treats a single label, 'strict liability', as though it named a single kind of offence, when it names a range of offences whose relationship to the principle of fault varies widely across the range.