The Wall Street Journal’s editorialists have rung the alarm bell on “The Threat of ‘Innovator Liability.’” They’re right to call attention to this latest weed in the liability landscape, but they’ve missed a big piece of the story.

Wyeth Inc. (a big pharmaceutical company) invented, patented, and until 2002 sold an acid reflux drug called Reglan. The drug has since been sold by a generic company. Danny Weeks claims to have suffered bad side effects from taking the generic stuff, which he insists was mislabeled. Naturally, he sued; naturally, for fraud and failure to warn; naturally, in state court (Alabama). Less naturally, he sued not the generic manufacturer but Wyeth.

Why would he (or rather his lawyers) do such a thing? Answer, because a federal statute commands generics to market and sell their products with the same label (meaning the incomprehensible info sheet inside the package) that is or was used by the patent manufacturer. By law, the generics have no control over the label. Therefore, the Supreme Court decided in Pliva v. Mensing (2011), the federal law preempts “failure to warn” suits against generics under state statutory or common law. (If federal law commands a private actor to do “A,” states can’t turn “A” into a tort or crime.) This is why Mr. Weeks’ lawyers made up the exotic “innovator liability” theory: it allows them to reach the company that invented the drug and marketed it a decade or so ago, under the offending label. In an 8-1 decision, the Alabama Supreme Court said: sure, go ahead.

Obviously, that’s nuts. What the WSJ doesn’t tell its readers (probably because it’s kind of complicated) is that the nuttiness has quite a bit to do with the Supreme Court—not Alabama’s, but the United States’.

Suppose Wyeth were still producing and selling Reglan alongside the generics: why is it possible at all to sue the patent manufacturer (and no one else) for “misbranding” and failure to warn? Because the Supreme Court said you can. The generic label must be identical with the patent producer’s label, which must be approved and is effectively dictated by the FDA—wording, font size, and all. However, said the Supreme Court in Wyeth v. Levine (2009), the elaborate FDA labeling regime does not preempt state law. (The theory is that the patent manufacturer can request the FDA to permit a label change. If the firm doesn’t and something bad happens, that’s “failure to warn.”) The decisive fifth vote for that holding came from Justice Thomas, along with a take-no-prisoners concurrence that contained very harsh words for the Supreme Court’s habit of finding preemption in statutes that (by Justice Thomas’s lights) compel no such result. Justice Thomas also wrote for the 5-4 majority in Pliva. There, he acknowledged that the result—liability for the patent folks but not for generics—made absolutely no sense. But that, he said, is for Congress to fix. No one here but us textualist chickens.

Wyeth, I’ve urged here  and elsewhere, is wrong for any number of reasons. Wyeth v. Weeks is simply an occasion to explain, one more time, the political economy of federal preemption:

“Let Congress fix it” is a counsel of despair. Congress can (on a very good day) “fix” this or that statutory detail, provided the folks who want it fixed pay up. But preemption provisions (express, or provisions bearing on the scope of preemption) aren’t details; they are the statute. They embody the basic compromise between producers who want federal protection and pro-regulatory constituencies who want to keep the state courthouse doors open. You can’t monkey with those provisions without re-opening the entire statutory bargain—which no one in his right mind wants, because anything could happen. Thus, either the Court makes sense of the statutes or else, it’s mayhem.

The flipside is this: unless you close absolutely every door by means of preemption, the trial lawyers will find an opening and in short order overrun the statute. Think “innovator liability” is far-fetched? Try this: generics can be held liable for “design defects” and “failure to warn” because they can comply with both the federal “same label” requirement and conflicting state duties by not selling the drug. Plaintiffs lawyers managed to sell a federal appeals court on this “stop selling liability” theory. The Supreme Court will review the decision. (Mutual Pharmaceutical Co. v. Bartlett, scheduled for oral argument on March 19.) The Court will reverse–

and the trial lawyers will crank up yet another theory. They are the most entrepreneurial profession in America; their returns on investment are very high; and some court somewhere in this vast country is always open for their business. Partial preemption is merely a temporary tax on their activity. What’s needed in this field is a prohibition against innovation—and a preemption doctrine that fits the bill.