Smilansky, in "Free Will, Fundamental Dualism, and the Centrality of Illusion" in Robert Kane’s excellent book, The Oxford Handbook on Free Will, argues that even though libertarian, contra-causal free will doesn’t exist, believing that it does is essential to maintain strong moral commitments, respect for human agents, and meaning in life. This is by far the most explicit argument mounted for a two-tiered system of knowledge and belief: there are those in the know (mostly philosophers and scientists) who realize the naturalistic truth about ourselves, and then there’s the rest of us - the vast majority who must be misled as to our real nature, lest we become demoralized. Because naturalism denies that we have free will, according to Smilansky it is too morally dangerous an idea to justify its dissemination. This suggests that Smilansky might oppose publication of such works as Owen Flanagan’s The Problem of the Soul (see section on Flanagan) and would advocate censorship of any discussion of naturalism. See my discussion of Smilansky in "Is Free Will a Necessary Fiction?", in which I strongly contest his claim that we need the illusion of free will to lead meaningful, moral lives, and that therefore we need not sequester knowledge about who and what we really are.
The Undiscovered Mind, 1999 Free Press, p. 247: "If free will is an illusion, however, it is an absolutely necessary one, more so than God." Horgan joins several commentators, including Stephen Pinker (How the Mind Works, see below), and Saul Smilansky (see above) who suppose that that we need the fiction of free will. (Most recently, see his New York Times piece on free will.) But how do we maintain such an illusion in the face of what is becoming ever more common knowledge about our causal antecedents in both genetics and the environment? Will those who champion illusionism, such as Horgan and Smilansky, actually seek to restrict understanding about naturalism and its viability as a world view? Have they really investigated the potential of naturalism to ground our legal and moral practices, and its implications for finding meaning in life? If not, then they are on rather shaky ground in opposing education about naturalism.
Matt Ridley, Genome
The last chapter is on free will: "Full responsibility for one’s actions is a necessary fiction without which the law would flounder, but it is a fiction all the same" (p.309). Actually, the law might not flounder, but retributive justifications for punishment might lose force. See the Criminal Justice page for extended discussion of this point, and see especially Derk Pereboom’s Living Without Free Will for an argument that we don’t need belief in free will to defend viable moral and criminal justice practices.
"Free will is an idealization of human beings that makes the ethics game playable. Euclidean geometry requires idealizations like infinite straight lines and perfect circles, and its deductions are sound and useful even though the world does not really have infinite straight lines or perfect circles. The world is close enough to the idealization that the theorems can usefully be applied. Similarly, ethical theory requires idealizations like free, sentient, rational, equivalent agents whose behavior is uncaused, and its conclusions can be sound and useful even though the world, as seen by science, does not really have uncaused events. As long as there is no outright coercion or gross malfunction of reasoning, the world is close enough to the idealization of free will that moral theory can meaningfully be applied to it" (p. 55). I take Pinker considerably to task for this in "Materialism and Morality: the Problem with Pinker".
Update: Pinker's newer book, The Blank Slate, revised his views on free will, in that he no longer thinks it's a necessary fiction. The chapter on "The Fear of Determinism" takes an explicitly deterministic stance, and usefully demonstrates the absurdity of contra-causal free will and why we shouldn't worry about being fully caused creatures. However, Pinker remains conservative in not drawing any conclusions about how not having free will might affect our attitudes towards punishment, credit, and blame,; that is, he doesn't explore the implications of determinism for ethical theory. This, despite the fact that in How the Mind Works he claimed that "ethical theory requires idealizations like free, sentient, rational, equivalent agents whose behavior is uncaused" (see above). We await further progress by Pinker.