Legal Commentary

San Francisco is now locked in a struggle with business over whether subnational governments can mandate that employers provide their employees with health care benefits. The employers are claiming that ERISA preempts the mandate, and their argument illustrates the insidiously anti-democratic nature of preemption arguments. As a matter of policy, I tend to agree that funding public benefits like health care through mandates on employers is foolish. Such a finance mechanism interferes with the mobility of labor and discourages job creation. Far better, it seems to me, to provide health benefits through general taxes not incident on employment.

But here is where I am a die-hard lover of federalism: As dumb as employer mandates are, centralizing debate over health care through a broad construction of ERISA preemption is even dumber. Such centralization is an outrage against the democratic process both locally (by suppressing the efforts of those zany San Franciscans) and nationally (by letting Congress off the hook of confronting the relationship between health care and employment). San Francisco hurts no one but itself and its own residents by burdening business and driving away capital to the 'burbs. The claim that national businesses will suffer some external cost outside San Francisco from disuniform regulation is patently baloney: Any business that operates in any city already must uncontroversially incur the costs of researching and complying with local zoning codes, local taxes and fees, local building codes, local safety regulations, etc. The marginal cost of insuring that one's local branch complies with the local complying health care law is close to zero.

Therefore, you can imagine my delight when I read Judge William Fletcher's opinion in Golden Gate Restaurant Association v. San Francisco, 546 F.3d 639 (9th Cir. Sept 30th 2008), upholding San Francisco's mandate on employers employing between 20-99 persons to provide health benefits to their employees. The Golden Gate Restaurant Association argued that ERISA preempted the ordinance because the ordinance "related to" or "made reference to" ERISA benefit plans. Fletcher rejected this argument, which had been adopted by the Fourth and Second Circuits, reasoning that San Francisco's ordinance did not regulate the terms of any employers' ERISA plans but rather required employers to provide a dollar amount of benefits (roughly $1.17-1.76 per employee hour) regardless of the terms of any employment contract.

Judge Fletcher is right on the legal merits, in my view, for reasons too tedious to detain anyone except aficionados of federalism and/or ERISA. (I include a few of those reasons after the jump for the one or two readers who might conceivably care about such matters). But, as far as "decentralization policy" is concerned, Fletcher's opinion is a precarious victory for sane federalism -- precarious, because it is vulnerable to en banc and SCOTUS review. One can only hope that SCOTUS's recent endorsement of the presumption against preemption in Altria v. Good will slow down the madness that is ERISA preemption.

For those who care about ERISA, why do I claim that preempting San Francisco's ordinance is madness? The Restaurant Association is essentially making an effects-based preemption argument, asserting that SF's ordinance effectively requires employers to change their ERISA benefits plans to comply with San Francisco law. The folly of this argument, however, is that it proves too much: Lots of local laws might have effects on employers' incentives to provide contractual benefits. Medical malpractice lawsuits under state tort law might drive up the cost of insurance, leading the marginal employer to reduce employees' health care benefits. Local zoning law could -- indeed, does -- increase housing costs, which increases the relative attractiveness of housing benefits to employers. But no lawyer in their right mind would argue that these state and local laws "relate to" ERISA benefits plan, because these laws' obligations are not triggered by the existence of ERISA-covered employment benefits.

It is conceded that SF could not impose mandates on employers because they provide ERISA-covered benefits to their employees. But SF has not done so: The duties under the SF ordinance apply to every employer, regardless of whether those employers provide ERISA benefits -- or any benefits - -pursuant to contract. True, employers who provide health care benefits thereby comply with the ordinance. But employers who do not do so do not thereby escape the ordinance -- and it is this latter fact that makes the ordinance a simple mandate on businesses rather than a regulation of the terms of ERISA-covered plans.

Any other theory will draw the courts into a theory of preemption that could suck every state and local regulation of business into the maw of ERISA preemption -- an outcome utterly unintended by anyone in Congress in the 1970s, when ERISA was enacted. For courts to create such centralization without Congress' assent is, as I noted above, an outrage against common sense and subnational democracy. As I have argued elsewhere (Against Preemption: How Federalism Can Improve the Federal Legislative

Process, 82 N.Y.U. L. Rev. 1 (2007)), ERISA preemption has also absolved Congress of the duty to confront the problem of how health care benefits relate to employment. Preemption, in short, destroys both subnational and national democracy. So here's hoping that Judge Fletcher can sustain his blow for federalism.