It should first be noted that three strikes and other mandatory minimum sentencing rules mark a shift in the role of the extrademocratic state. While libertarians bemoan the intrusive role of the welfare state in the lives of a free citizenry, they are remarkably silent with regard to the role of criminal law enforcement, which they regard as necessary to the protection of private property and bodily security.⁵  Yet the police apparati—patrol officers, prosecutors, the judiciary, jailers—represent one of the greatest threats to a free citizenry, a delicate pragmatic compromise that a democracy must make to accommodate the day-to-day necessities of social order. J. Edgar Hoover's extortionist and informationalist FBI, and more recently, Darryl Gates's defiant attitude toward civilian supervision of the L.A. police force, demonstrate what can happen if those entrusted with the duty of enforcing society's priorities aspire to set those priorities. The empirical presuppositions and practical ramifications of three strikes–type legal rules each ensure an intensified role for an increasingly powerful police bureaucracy, a full-fledged "fifth estate" whose authority will be difficult to check and whose influence already dwarfs its limited mandate to protect and serve. 

Criminal law has always served a number of functions simultaneously. Michel Foucault understood that the most important function of criminal law is not how it affects the criminal (punishment or therapy) or even the potential criminal (deterrence), but how it affects, by defining, the law-abiding citizen: criminal law definesthe boundaries of acceptable conduct in a society, and in so doing it attempts to give a society an identity. The law is in this sense both a regulative and an epistemological technology: it generates the rationale for the monitoring of citizens (the prevention of crime) while its rules and decisions produce norms of behavior, the texts in which society writes, reads—and therefore knows—itself. Thus while proponents of draconian sentencing provisions inevitably claim that the new laws will "send a message to criminals," the real recipient of the message is society itself. 

It is useful to examine the empirical claims of three strikes—claims one must accept in order to view three strikes as rational policy rather than reactionary demagoguery. Three strikes proponents claim that most violent crime is committed by a small but hard-core group of criminal recidivists (some claim 5 percent of the "criminal population") who should be permanently isolated from law-abiding society. This is an attractive notion because it suggests that American society and Americans in general are not responsible for the violent character of American life; instead it is a few "bad seeds" who are spoiling the whole culture. The "three-time loser" of the crime bills becomes the born loser who cannot be reformed, but rather must be removed from society. Three strikes tells us that the inchoate sense of insecurity and alienation many Americans feel is attributable to a small minority of wrongdoers who can be removed from society at little cost. 

The truth of America's crime wave is, literally, elsewhere. Even as the public perception of jeopardy has risen, actual crime statistics show a slight decrease in crime.⁶  People at the least risk of becoming victims of crime fear crime the most, demonstrating that fear of crime is a function of something other than actual peril. Moreover, violent crime (one category that has shown a slight increase) is often perpetrated by someone familiar to the victim—a friend, lover, fellow employee, or family member. The "bad seed" theory of criminal recidivism simply doesn't hold up under scrutiny, and although there are of course recidivists, they are not primarily responsible for the type of crime the public fears most. 

It was only a matter of time before the "therapeutic state" began to influence criminal justice. Three strikes and the theory of recidivist-driven crime that underlies it represent criminal law as public therapy. Three strikes makes many feel good, both because as a society we are doing something about crime and because we need not confront the possibility that pervasive social violence and the perception of rising crime may be a function of pervasive social practices that nonrecidivists also engage in. Good therapy may mask bad policy. So the carceral state looms behind the soothing but impotent figure of the therapeutic state, the latter the lure that serves the former like the song serves the siren.