The real problem is that litigation is an extremely expensive and long-lead-time method of regulation. Very few trial court decisions function effectively as regulation—at least not by themselves. A result significant enough to provide a motive to behave as if regulated will almost always be appealed at least once. That means time, money, and a crushing emotional burden on the parties, not to mention the opportunity cost of having their attention diverted from better activities.
In a sense, though, litigation-as-regulation is a reaction to the governmental equivalent of a "market failure." Just as what regulation is supposed to do is deal with instances in which private parties in general cannot behave "equitably" in the aggregate—usually from asymmetrical information, grossly unequal bargaining positions, or both—the same forces that cause the market to fail can also coopt regulatory agencies and efforts. Exhibit A: Drilling for oil in Artic preserves. Whether this is a wise policy or not is beside the point; only a moron could legitimately claim that the process has been transparent and fair.
Litigation, at least in the US, must come from a live dispute. It follows, therefore, that it is retrospective in nature, even when announcing a prospective rule (e.g., Miranda). Regulation, however, need not be retrospective; it can instead anticipate needs based upon potentials. Some would argue that markets are also prospective, based on the "discount to present value" theorem. This, however, is a mirage in practice, because markets can discount only that information known to all market participants. The reality is that market "pricing" is selectively retrospective, because the information necessary for all market participants to make rational decisions is not available to all of those participants at the same time. So, then, one might say that the reality of litigation and markets use inductive logic—reasoning from specific past instances to rules of general applicability—while the reality of regulation uses deductive logic—reasoning from general principles to specific prospective instances.
Copyright infringement illuminates these disjunctures. On the one hand, we have the Constitutional premise that "more Art is Good." The regulatory mechanism that has been chosen is a limited monopoly for creators—an interference in the marketplace. Litigation accusing others of violating those regulations, though, seldom leads to clear guidelines that are useful to the actual market participants. It is one thing entirely to say "parody may be fair use, and must be judged under the § 107 factors" (Campbell v. Acuff-Rose Music, Inc.); it is entirely another to look at a particular artistic effort that might be a parody (whether claimed so ab initio or in response to accusations of infringement is irrelevant to this inquiry) and determine whether for purposes of § 107 it is a parody. We have jumped between deductive and inductive reasoning at a dizzying pace—and we haven't even gotten to warranty and indemnity clauses yet…