Cities as Constitutional Actors: The Case of Same-Sex Marriage

University of Virginia Law School

Public Law and Legal Theory Working Paper Series

Year  Paper 

Cities as Constitutional Actors: The Case of

Same-Sex Marriage

Richard Schragger

∗

University of Virginia School of Law, [email protected]

This working paper is hosted by The Berkeley Electronic Press (bepress) and may not be commer-

cially reproduced without the permission of the copyright holder.

http://law.bepress.com/uvalwps/uva publiclaw/art30

2005 by the author.

Cities as Constitutional Actors: The Case of

Same-Sex Marriage

Richard Schragger

Abstract

In February 2004, the City of San Francisco began to issue marriage licenses to

same-sex couples; eventually the City would issue over 3,500 same-sex licenses.

Following San Francisco’s lead, other municipalities in New York, New Jersey,

Oregon, and elsewhere also began to issue marriage licenses to same-sex couples.

These municipal actions fueled an already bitter national debate over the morality

and legality of same-sex marriage. This Essay examines the legal and constitu-

tional role of local governments in this debate. I make two claims. First, I argue

that, despite the well-publicized actions of local ofﬁcials in issuing same-sex mar-

riage licenses, scholars have generally overlooked the possibility and desirability

of local determinations of marriage eligibility. Conventional wisdom, based on

long-standing tradition, has assumed that marital status (and domestic relations

law in general) should remain the province of the states. But few have asked

the functional question: At what level of government should marriage eligibil-

ity criteria be determined? I claim that there is no reason that local governments

cannot be tasked with the power to make marriage eligibility determinations, and

indeed, there are many good reasons for them to do so. Second, I argue, more

provocatively, that the Court’s equal protection doctrine might require that local

governments be permitted to make marriage eligibility determinations, at least

with regard to gays and lesbians. This argument is based on a “localist” reading

of the Supreme Court’s decision in Romer v. Evans. Romer, the Essay argues,

may require that state same-sex marriage bans be struck down insofar as they pre-

empt local decisions to recognize same-sex unions. The Essay then suggests the

contours of an equal protection doctrine that provides a realm of “constitutional

home rule” by protecting local governments from contrary state commands in the

course of vindicating substantive constitutional rights. My account of “localist

constitutionalism” asserts that local governments are importantly different from

states in a number of ways that might be salient in determining the content of con-

stitutional rights. This may mean that under some circumstances, localities should

be permitted to regulate in areas that states cannot, free from state interference. In

the context of same-sex marriage this “decentralized equal protection” jurispru-

dence would prevent states from interfering with a local government’s decision to

marry gay and lesbian couples.

Cities As Constitutional Actors: The Case of Same-Sex Marriage

Working Paper: June 7 Draft

Richard C. Schragger



In February 2004, the City of San Francisco began to issue marriage licenses to same-sex couples;

eventually the City would issue over 3,500 same-sex licenses. Following San Francisco's lead, other

municipalities in New York, New Jersey, Oregon, and elsewhere also began to issue marriage licenses to

same-sex couples. These municipal actions fueled an already bitter national debate over the morality and

legality of same-sex marriage. This Essay examines the legal and constitutional role of local governments

in this debate. I make two claims. First, I argue that, despite the well-publicized actions of local officials

in issuing same-sex marriage licenses, scholars have generally overlooked the possibility and desirability

of local determinations of marriage eligibility. Conventional wisdom, based on long-standing tradition,

has assumed that marital status (and domestic relations law in general) should remain the province of the

states. But few have asked the functional question: At what level of government should marriage eligibility

criteria be determined? I claim that there is no reason that local governments cannot be tasked with the

power to make marriage eligibility determinations, and indeed, there are many good reasons for them to do

so. Second, I argue, more provocatively, that the Court's equal protection doctrine might require that local

governments be permitted to make marriage eligibility determinations, at least with regard to gays and

lesbians. This argument is based on a “localist” reading of the Supreme Court’s decision in Romer v.

Evans. Romer, the Essay argues, may require that state same-sex marriage bans be struck down insofar as

they preempt local decisions to recognize same-sex unions. The Essay then suggests the contours of an

equal protection doctrine that provides a realm of “constitutional home rule” by protecting local

governments from contrary state commands in the course of vindicating substantive constitutional rights.

My account of “localist constitutionalism” asserts that local governments are importantly different from

states in a number of ways that might be salient in determining the content of constitutional rights. This

may mean that under some circumstances, localities should be permitted to regulate in areas that states

cannot, free from state interference. In the context of same-sex marriage this “decentralized equal

protection” jurisprudence would prevent states from interfering with a local government’s decision to

marry gay and lesbian couples.

Introduction

On February 10, 2004, the Mayor of the City and County of San Francisco, Gavin

Newsom, wrote a letter to the Director of the County Clerk’s Office, Nancy Alfaro, asking her to

“determine what changes should be made to the forms and documents used to apply for and issue

marriage licenses in order to provide marriage licenses on a non-discriminatory basis, without

regard to gender or sexual orientation.”

Following the Mayor’s instructions, the Director

created a “gender-neutral application for public marriage licenses and a gender-neutral marriage



Associate Professor, University of Virginia School of Law. Many thanks to John Harrison, Roderick Hills, David Barron, Brian

Bix, Ariela Dubler, Mark Rosen, Laurie Reynolds, Carlos Ball, Elizabeth Scott, Clay Gillette, Kim Forde-Mazrui, and

particularly, Risa Goluboff, for helpful comments and conversations. A preliminary version of this paper was presented at the

Virginia Journal of Law and Politics’ symposium entitled Democracy in Action? The Law and Politics of Local Governance,

and at a Virginia Summer Faculty Workshop.

Lockyer v. City and County of San Francisco, 33 Cal. 4th 1055, 1069-1070 (2004) (quoting Letter from Gavin Newsom, Mayor

of San Francisco, to Nancy Alfaro, Director of the County Clerk’s Office (Feb. 10, 2004).

Hosted by The Berkeley Electronic Press

license.”

Two days later, the county clerk’s office began issuing marriage licenses to same-sex

couples and the county recorder registered marriage certificates submitted on behalf of same-sex

couples who had received marriage licenses from the city and had participated in marriage

ceremonies. Over the next four weeks, the city issued over 3,500 marriage licenses to same-sex

couples.

A number of other municipalities throughout the country soon followed San Francisco’s

example. The clerk of Sandoval County, New Mexico began issuing same-sex marriage licenses

on February 20 before being ordered to stop by a state court.

In Oregon, Multnomah County

issued over 3,000 same-sex marriage licenses before being ordered to stop;

the county

commissioners of Benton County voted to stop issuing all marriage licenses rather than issue

only opposite-sex licenses.

In Asbury Park, New Jersey, the local registrar issued more than a

dozen same-sex marriage licenses before she was threatened with prosecution by the state

attorney general; in response the city filed a lawsuit against the state and the attorney general.

In New York, the City of Nyack initiated a suit against the state and the town clerk of

Orangetown for the right to issue same-sex marriage licenses,

the mayor of New Paltz

performed a number of same-sex marriage ceremonies for which he was prosecuted by a local

Id.

Id. at 1071.

Evelyn Nieves, Calif. Judge Won’t Halt Gay Nuptials; New Mexico County Briefly Follows San Francisco’s Lead; WASH.

OST, Feb. 21, 2004, at A1. Sixty-six marriage licenses were issued by County Clerk Victoria Dunlap before New Mexico

Attorney General Patricia Madrid’s ordered a halt. Id. A March 2004 restraining order against Dunlap was later upheld by the

Supreme Court of New Mexico, without comment, on July 8. Michael Davis, Dunlap’s Same-Sex Marriage Suit Stalls,

LBEQUERQUE JOURNAL, July 9, 2004, at B2. Litigation was dismissed in January 2005, shortly after the expiration of Dunlap’s

term in office. Joshua Akers, Lawsuit Against Dunlap Dismissed, A

LBEQUERQUE JOURNAL, January 4, 2005, at D1.

Sarah Kershaw, Oregon Supreme Court Invalidates Same-Sex Marriages, N.Y. TIMES, April 15, 2005, at A12. Following the

passage of an Oregon state constitutional referendum in November limiting marriage to opposite-sex couples, the Oregon

Supreme Court ruled all the Multnomah County same-sex marriages unlawful and invalid. Li v. Oregon, 110 P.3d 91 (Ore.

2005).

Kate Zernike, Gay? No Marriage License Here. Straight? Ditto., N.Y. TIMES, Mar. 27, 2004, at A8.

Thomas Crampton, Asbury Park Halts Gay- Marriage Applications, Sending Issue to Courts, N.Y. TIMES, Mar. 11, 2004, at B5.

Thomas Crampton, In a Lawsuit, Same-Sex Couples Say New York State Ruined Their Wedding Plans, N.Y. TIMES, Apr 8,

2004, at B4.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

district attorney,

and the mayor of Ithaca announced that she would accept marriage license

applications from same-sex couples and forward them to the state health department.

Massachusetts, where the Supreme Judicial Court has ordered the recognition of same-sex

marriages, a number of Massachusetts towns, including Plymouth, Attleboro, Fall River,

Provincetown, Somerville, Springfield, and Worcester, issued marriage licenses to out-of-state

couples who intended to remain living out-of-state despite the attorney general’s order that they

not do so.

In all these instances, local government officials—namely city mayors and county

clerks—interpreted state statutes to allow the issuance of same-sex marriage licenses, or read

state constitutional or federal constitutional guarantees of equal treatment as overriding state

laws that barred such licenses. Their actions sometimes put them directly in conflict with

unambiguous state law, and almost always put them in conflict with state officials.

Local officials’ actions also added fuel to a national, constitutional and political firestorm

that had been simmering since courts in Hawaii and Vermont ruled that same- sex couples had to

be granted marriage-like benefits,

and the Massachusetts Supreme Judicial Court ruled that

marriage itself had to be available to same-sex couples under state guarantees of equal

treatment.

In part goaded by the issuance of same-sex marriage licenses across the country,

President Bush backed a federal constitutional amendment that would ban gay marriage and, in

Id. Thomas J. Lueck, Police Charge New Paltz Mayor For Marrying Same-Sex Couples, N.Y. TIMES, March 3, 2004, at B4. A

judge later ruled against thirteen same-sex couples whose pending wedding plans were interrupted by the injunction against New

Paltz Mayor Jason West. Thomas J. Lueck, State Justice Rules Against 13 Couples Seeking Same-Sex Marriage, N. Y. Times,

Dec. 8, 2004, at B4.

Thomas Crampton & Michelle York, Hoping Courts Will Address Same -Sex Marriage, Ithaca Begins Accepting Licenses,

N.Y. T

IMES, March 2, 2004, at B4.

Cote-Whitacre v. Dep’t of Public Health, 2004 WL 2075557, at *5 (Mass. Super. Ct. 2004; Tamara Race, Plymouth Accepts

Out-of-State Gays’ Residency Oaths, PATRIOT LEDGER, May 27, 2004). In Takoma Park, Maryland, the city council issued a

resolution supporting same-sex marriage, though the city stated that it could not issue marriage licenses. Tarron Lively, Takoma

Park Backs Gay Unions, W

ASH. TIMES, July 15, 2004, at B3.

Baehr v. Lewin, 852 P.2d 44 (Haw. 1993); Baker v. State, 744 A.2d 864 (Vt. 1999).

Goodridge v. Dep’t of Public Health, 798 N.E. 2d 941 (Mass. 2003); Opinions of the Justices to the Senate, 802 N.E.2d 565

(Mass. 2004).

Hosted by The Berkeley Electronic Press

November, eleven states adopted through referendum their own state-level constitutional bans on

same-sex marriage.

The relationship between national, state, and local power, and the

appropriate level of government at which to regulate marriage has become intimately tied up

with the substantive questions of the morality of homosexuality and the justice of same-sex

marriage.

This Essay examines the legal and constitutional role of local governments in the on-

going debate about same-sex marriage. Many are predicting that the Court will eventually have

to determine the question of whether excluding gays and lesbians from state-sanctioned marriage

violates the Equal Protection or Due Process Clauses. These commentators point to the Supreme

Court’s recent decision in Lawrence v. Texas,

which struck down a state sodomy statute that

applied only to homosexuals. In Lawrence, the Court held that the fundamental right of intimate

association extended to homosexual sexual relations. Commentators also point to the Court’s

1996 decision in Romer v. Evans,

in which the Court struck down a Colorado state

constitutional amendment that barred the state, its agencies or local governments from adopting

anti-discrimination ordinances to protect gays and lesbians.

For some, Lawrence and Romer indicate that a ruling striking down bans on same-sex

marriage is only a matter of time.

That optimism may be premature, however. In light of the

overwhelming success of same- sex marriage bans in the states and George W. Bush’s reelection,

a Court attentive to popular opinion might think twice about overturning traditional heterosexual

The eleven states were Arkansas, Georgia, Kentucky, Michigan, Mississippi, Montana, North Dakota, Ohio, Oklahoma,

Oregon, and Utah. Some states had already adopted such amendments before November 2004. See, e.g., A

LASKA CONST. art. I,

§ 25 (adopted in 1998); H

AW. CONST. art. I, § 23; (adopted in 1998). Many states have adopted statutory bars to same-sex

marriage or to the recognition of out-of-state same-sex marriages. See William C. Duncan, Revisiting State Marriage

Recognition Provisions, 38 C

REIGHTON L. REV. 233 (2005).

539 U.S. 558 (2003).

517 U.S. 620 (1996).

See, e.g., Lawrence Tribe, Lawrence v. Texas: The “Fundamental Right” that Dare Not Speak its Name, 117 HARV. L. REV.

1893, 1945-47 (2004).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

marriage.

The Court may instead defer to the states on the matter, a position advocated by

many scholars and policy-makers on both sides of the marriage debate.

The conflict between states and localities over marriage, however, points to yet another

possibility—the potential that cities could be empowered to make decisions about the extent to

which gay and lesbian couples receive the benefits of marriage. This Essay examines the

doctrinal potential for a constitutional holding that makes it possible for cities like San Francisco

and other local governments to regulate marriage eligibility as a matter of local concern. I

contend that there is no reason that the eligibility requirements for marriage cannot be fixed at

the local level and that the constitutional requirement of equal protection may require that it be

so.

Part I takes up the first claim. I argue that, despite the well-publicized actions of local

officials in issuing same-sex marriage licenses, scholars have generally overlooked the

possibility and desirability of local determinations of marriage eligibility.

Conventional

wisdom, based on long-standing tradition, has assumed that marital status (and domestic

relations law in general) should remain the province of the states. To the extent that

conventional understanding is under pressure, it is from proponents of a national constitutional

amendment defining marriage as a union between a man and woman.

But while commentators

See Michael J. Klarman, Brown and Lawrence 29-30 (Jan. 6, 2005) (unpublished manuscript, on file with author) (arguing that

the Supreme Court is not likely to strike down bans on same-sex marriage when such bans are strongly supported by public

opinion).

Compare John C. Yoo & Anntim Vulchev, The Conservative Case Against the Federal Marriage Amendment, 2004 Issues in

Legal Scholarship, available at http://www.bepress.com/ils/iss5/art3/ with Franklin Foer, The Joy of Federalism, N.Y. T

IMES,

March 6, 2005, at 7 (noting that liberal policy-makers like Massachusetts Democrat Barney Frank have recently espoused

federalism and states-rights arguments in the gay marriage debate).

Jennifer Gerarda Brown mentions the possibility of local recognition in passing in her article, Competitive Federalism and the

Legislative Incentives to Recognize Same-Sex Marriage, 68 S. C

AL. L. REV. 745, 809 (1995), but her article assumes state control

of marital status. See also Brian H. Bix, State Interests in Marriage, Interstate Recognition, and Choice of Law, 38 CREIGHTON

L. REV. 337, 338-39 (2005) (noting that local recognition makes sense as a policy matter, but that American family law has

historically been under state control).

See Edward Stein, Past and Present Proposed Amendments to the United States Constitution Regarding Marriage, 2004 Issues

in Legal Scholarship; see also Defense of Marriage Act, 1 U.S.C. § 7, 28 U.S.C. 1783C (2005); Marriage Protection Amendment,

S.J. Res. 1, 109th Cong. (2005).

Hosted by The Berkeley Electronic Press

have begun to debate the appropriateness of a national marriage law, they have mostly failed to

ask the functional question—at what level of government should marriage eligibility criteria be

determined? I argue that there is no reason that local governments cannot be tasked with the

power to make eligibility determinations, and indeed, there are many good reasons for them to

do so.

Part II then makes the more provocative assertion, arguing that the Constitution might

require that local governments be permitted to make marriage eligibility determinations, at least

with regard to gays and lesbians. This argument assumes that the Court is not going to mandate

same-sex marriage in the near future through some other means, for example, by recognizing

homosexuals as a suspect class under equal protection analysis. It thus employs existing

precedent, namely Romer v. Evans, to make the case for a local option. Romer is a notoriously

opaque decision, but it has been read by some commentators as preserving a sphere of local

authority to adopt policies that might be contrary to state preferences under certain

circumstances.

This “localist” reading of Romer is directly relevant to the constitutionality of

same-sex marriage bans in the states. I argue that Romer may require that state same-sex

marriage bans be struck down insofar as they preempt local decisions to recognize same-sex

unions.

Finally, Part III suggests the contours of a constitutional doctrine that provides a realm of

“constitutional home rule” by protecting local governments from contrary state commands in the

course of vindicating substantive constitutional rights. My goal is to describe in the broadest

terms how an equal protection doctrine attentive to the scale of government action could create a

truly devolutionary constitutional jurisprudence. Romer gestures toward such a jurisprudence,

See, e.g., David Barron, The Promise of Cooley’s City: Traces of Local Constitutionalism, 147 U. Pa. L.. Rev. 487, 586-94

(1999); Lawrence Rosenthal, Romer v. Evans and the Transformation of Local Government Law, 31 Urban Lawyer 257 (1999).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

but only slightly, and any further movement in that direction would require the Court to jettison

two deeply held constitutional conventions. The first convention is that rights are fixed across

levels of government. The second convention is that states exercise plenary power over their

local governments. My account of “localist constitutionalism”

asserts that local governments

are importantly different from states in a number of ways that might be salient in determining the

content of constitutional rights. This may mean that under some circumstances, localities should

be permitted to regulate in areas that states cannot, free from state interference. The result, as

applied in the context of same-sex marriage, is that states cannot override a decision by a local

government to recognize same-sex marriages.

Should Local Governments Determine the Eligibility Criteria For Marriage?

The issuance of same-sex marriage licenses by cities raises an obvious question, though

one that many scholars would dismiss out-of-hand: Should local governments be able to decide

who is eligible to be married? The conventional wisdom has been that family law, including

marital status, is the province of the state.

The idea that numerous local governments could

decide who is eligible to be married would strike many as absurd.

Yet it is not clear why this is so. As one commentator has observed, the fact that states

are charged with regulating the family “might be seen as . . . an accident of history . . . it would

not have been entirely unworkable or contrary to the express language of the Constitution for

domestic regulation to have been centered primarily at the national level or dispersed to the local

See Barron, supra note 22; Rosenthal, supra note 22.

It should be noted that the federal government has always been involved, and increasingly so, in the regulation of families. See

generally, Ankenbrandt v. Richards, 504 U.S. 689 (1992); Parental Kidnapping Prevention Act, 28 U.S.C. § 1738A (2003);

Defense of Marriage Act of 1996, Pub. L. 104-199 (codified as amended in 1 U.S.C. § 7 and 28 U.S.C. §1738C); Mail-Order

Brides Act of 1996, 8 U.S.C.A. § 1375 (West 1999); Interethnic Adoption, 42 U.S.C.A § 1996b (West 1994 & Supp. 1999);

Child Support and Recovery Act of 1992, Pub. L. No. 102-521 (codified as amended in 18 U.S.C. and 42 U.S.C.). Nevertheless,

family law generally remains largely state-based. See Sosna v. Iowa, 419 U.S. 393, 404 (1975) (stating that “[d]omestic relations,

[is] an area that has long been regarded as a virtually exclusive province of the States.”).

Hosted by The Berkeley Electronic Press

level.”

The assertion that states should make the rules about marriage assumes and celebrates a

significant amount of diversity across jurisdictions, at least as a theoretical matter. This

commitment to state-level diversity is itself being challenged by those who favor a national

marriage amendment. Though there has been significant opposition to the amendment on

federalism grounds—again, the conventional wisdom being that states should be the site of

marriage regulation

—there has been little engagement with the functional question of where

marriage eligibility is best determined. The assumption that marriage eligibility needs to be

uniform across a state, but not across the nation, seems mostly based on tradition and a vague

commitment to federalism, and even those that advocate national rulemaking rarely ask

functional questions about the appropriate site for governmental regulation. Few inquire whether

states are the appropriate units for determinations of this kind or whether some other unit of

government would be equally as, or more, appropriate.

My argument here is that local determinations of marriage eligibility should not strike us

as odd to the extent that we support some level of diversity of marriage eligibility regimes.

Certainly, the general arguments in favor of the current state-centered marriage regime can be

marshaled in support of additional decentralization. Federalism is often justified as a mechanism

for encouraging state-level experiments, increasing political accountability, promoting inter-state

competition and innovation, and developing citizens’ skills of self-government. There is nothing

magical about states, however. I and others have argued that states are often too large to achieve

the oft-asserted goals of federalism, and that those ends would be better served by devolving

power to localities.

The principle of subsidiarity – which asserts that to the extent possible

Bix, supra note 20, at 339.

See Yoo & Vulchev, supra note 19.

See Jerry Frug, Decentering Decentralization, 60 U. CHI. L. REV. 253 (1993); Richard C. Schragger, Reclaiming the

Canvassing Board: Bush v. Gore and the Political Currency of Local Government, 50 BUFF. L. REV. 393, 424 (2002) [hereinafter

“Reclaiming Canvassing Board”].

http://law.bepress.com/uvalwps/uva_publiclaw/art30

decisions should be made at the most local level – supports a presumption of local decision-

making.

Indeed, local determinations of marriage eligibility seem most consonant with our

intuitions about the purpose of the public solemnization of marriage. To the extent that public

recognition reflects the idea (or at least our idealization) of marriage as a public enterprise, a

statement of mutual aid and respect that is made meaningful by its sanctioning in a particular

community, the relevant community seems to be the local one. The idea that the national or state

political communities are interested actors in such a small-scale public commitment seems

contrary to the intimacy of the sanctioning exercise. In a balancing of relevant governmental

interests, local citizens appear to have a greater interest in who among them is able to access the

benefits of marriage than do citizens who live outside the jurisdiction. The local jurisdiction is

charged with the kind of quotidian regulation that affects the health and welfare of its citizens

most directly, in particular the welfare of its families. Local governments have a great deal of

interest in the formation of the families that will reside in the jurisdiction, the stability of family

units, and the rules that provide for such formation.

A. Administrative Efficiency

Nevertheless, marital status is governed by state statute in all states, and even in those

states with robust constitutional or statutory grants of local home rule, domestic relations has

fallen within what commentators and courts call the “private law exception” to home rule

authority.

The “private law exception” presumes that a whole range of regulatory activities are

inappropriate for local determination, including such common law subjects as torts, contract, and

property, as well as domestic relations law, including marital status.

The rationale for this

See Gary Schwartz, The Logic of Home Rule and the Private Law Exception, 20 UCLA L. REV. 670 (1973).

See id.; see also RICHARD BRIFFAULT & LAURIE REYNOLDS, STATE AND LOCAL GOVERNMENT LAW 309-12 (6th. ed. 2001).

Hosted by The Berkeley Electronic Press

exception to home rule grants of authority appears to be efficiency. The argument is that a

multiplicity of local contract, tort, domestic relations, or property laws would generate extreme

inefficiencies both in terms of information costs and negative spillovers.

But efficiency concerns seem overblown with regard to the local determination of who is

eligible to be married. First, marriage is entered into voluntarily; it is a status that does not

normally come as a surprise, as incidental to some transaction of business or otherwise day-to-

day activity. Because marriage is a status, the requirements to enter into it will not normally

alter primary conduct, such as one’s standard of care or one’s duty to customers or contracting

partners. Second, it is easy to ascertain the local rules for entering into marriage—one need only

examine the relevant forms or contact the jurisdiction’s relevant authorities. Parties entering into

the marriage relationship currently have to comply with state statutes (often directing them to

apply for licenses with local government officials); they can easily comply with equivalent local

ordinances. Finally, it is relatively easy for courts to determine one’s status if disputes should

arise. All a court needs to know is whether the locality has issued a license and the jurisdictional

reach of that government’s authority. Indeed, numerous municipalities have already adopted

domestic partnership ordinances that provide rules for the recognition and dissolution of

marriage-like partnerships. Though these ordinances cannot require the state or private parties to

treat domestic partners as married for all purposes, a number of ordinances mandate enforceable

duties and obligations on the part of the partners entering into the relationship or extend spousal

benefits to partners of municipal employees. Other municipal ordinances require private actors

that contract with the city to extend domestic partnership benefits to their employees.

See generally Chad Bayse, Minneapolis Wades into Domestic Partner Benefits Legislation Once Again, 30 William Mitchell L.

Rev. 931 (2004).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

One may object that intra- and inter-state determinations of who is and who is not

married might become complicated. But why? Imagine a system in which cities are empowered

to allow same-sex couples who live within their borders to opt into the state’s marriage regime.

The determination of status would be geographical: while the couple was domiciled in San

Francisco, they would be married according to the laws of the City and County of San Francisco

and the State of California. If they moved out of San Francisco or California to a jurisdiction

that does not recognize same-sex marriage, their legal status would change from married to non-

married.

A forum-selection clause could ensure that status be determined based on the law of

the city of celebration, as long as one party continued to reside there.

Of course, marital status implicates a whole range of legal rights; the state would have to

determine how those rights change when residence changes. Local cross-border moves from

jurisdictions that recognize same-sex marriage to jurisdictions that do not would require state-

level rules regarding how such residency changes affect parental rights and child custody, tax

filing status, property division on dissolution, and many other legal rights and obligations. But

this is no different than the rules currently required under a federal regime that has significant

state-level diversity;

the issues are not unique to a system of local regulation. The difference is

The opt-in regime operates similarly to local domestic partnership ordinances, which permit localities to grant spousal-like

benefits to city workers and their dependents. The substantive benefits law does not change, only those eligible for benefits.

I am assuming for these purposes that neighboring local and state governments would not be required to recognize out-of-state

or out-of-locality marriages when the couples relocated to those jurisdictions. See infra note 44. Recognition, however, comes in

many shapes and sizes. One can imagine a rule whereby one’s same-sex marriage is only recognized as long as one is physically

within the borders of the locality, or one can imagine a rule whereby one’s same-sex marriage is recognized state-wide as long as

one is domiciled in the status-conferring jurisdiction. The latter rule makes more practical sense as inter-local travel is quite

commonplace, but it would admittedly have larger cross-border and third-party effects. See TAN infra at —

See e.g., ATLANTA, GA., CODE OF ORDINACES ch. 94, art. VII, § 94-137 (2000). Of course, courts will be faced with local-level

choice of law issues that parallel the state-level choice of law issues that already exist. In a unified state court system, this should

not be difficult; the state court will simply apply the relevant local marital law as it is dictated by domicile. Inter-state choice of

law issues will become somewhat more complicated, but only insofar as courts will have to determine the content of foreign local

law as opposed to foreign state law. To the extent that a state currently rejects application of out-of-state marital eligibility law

based on public policy, its courts will continue to do so whether that out-of-state law is state or municipal.

Parental Kidnapping Prevention Act of 1980, Pub. L. No. 96-611 (codified as amended in 28 U.S.C. and 42 U.S.C.); Uniform

Child Custody Jurisdiction and Enforcement Act (1997), 9 U.L.A. 257 (Supp. 1999); Uniform Child Custody Jurisdiction Act

(1968), 9 U.L.A. 115 (1988); Uniform Parentage Act (1973) 9B U.L.A. 369 (1987).

Hosted by The Berkeley Electronic Press

one of degree, not kind, and those differences may not be very significant in an era where family

status changes and mobility across both state and local jurisdictional lines is taken for granted.

In fact, because states have uniform court systems, states are in a better position to

accommodate intra-state, local changes in marital status than a federal system is in

accommodating similar inter-state changes. No doubt there are some additional costs of

administration in determining the appropriate rules for solving local cross-border questions, but

regulatory diversity is not unworkable merely because it involves local rather than state

regulators. The current inter-state settlements of these kinds of cross-jurisdictional questions

–

and admittedly a number of questions remain unsettled

– would remain unchanged.

Obviously, residence-based status changes introduce uncertainty both for the couples

involved and for third parties. From the couples’ perspective and from the perspective of a

central planner, a national rule that guaranteed a stable status would always be preferable. But

that is a problem for any sub-federal regime, not a particular drawback of a local one. As an

administrative matter, local residence-based status changes are not so different from many of the

other kinds of regulations with which one has to comply when one enters or leaves a jurisdiction.

Private actors already must adapt their conduct to account for jurisdictional diversity when they

pay taxes, sell goods, or seek licenses to do business or drive a car in a new jurisdiction. Perhaps

marital status feels different because it seems strange that the definition of one’s intimate

relationship could change with the crossing of a jurisdictional boundary. Indeed, the sense that

marriage is “special” seems to drive both the quest to gain—and to deny—access to it. From a

See e.g., Parental Kidnapping Prevention Act of 1980; Uniform Child Custody Jurisdiction and Enforcement Act (1997);

Uniform Child Custody Jurisdiction Act (1968); Uniform Parentage Act (1973).

In one case, a Vermont family court judge issued child visitation rights to a nonbiological parent upon the dissolution of her

civil union with the child’s biological mother only to have a Virginia court later declare that under that state’s “Affirmation of

Marriage Act,” parentage between same-sex partners is not recognized. The Vermont court subsequently reasserted jurisdiction

in an ongoing legal battle of the very sort avoided by federal and uniform laws for heterosexual couples. Jonathan Finer, Court

Says Both in Gay Union Are Parents, W

ASH. POST, November 22, 2004, at A3; see also Finstuen v. Edmondson (W.D. Ok. filed

Sept. 15, 2004).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

regulatory perspective, however, there are no compelling reasons why residence-based status

changes cannot be accommodated by the existing judicial and administrative system. If

increased administrative costs alone could overwhelm the argument for jurisdictional diversity,

then such costs would necessarily require national regulation across a wide spectrum of

activities.

B. External Effects

Administrative costs are one concern in determining the appropriate scale for government

action. Another is whether local actions have external effects. When deciding whether a local

regulation is of “statewide” or “local” concern, state courts making state-level home rule

determinations normally focus on the external effects of local regulation. State uniformity is

often asserted to prevent negative spillovers.

In this case, an opponent of local eligibility rules

would argue that the cross-border costs of local rules are too high.

An increase in the diversity of regulatory regimes always increases some costs. The

question here is whether those increased costs are significant, either in terms of state-wide effects

or specific cross-jurisdictional effects. As to the first, local marriage eligibility requirements will

affect the state’s domestic relations laws insofar as more or different persons will be eligible to

access the benefits and burdens of marriage. The substantive domestic relations law—taxation

of marital assets, property distribution on dissolution and death, marital rights vis-à-vis third

parties and all the rest—will remain the same, however. Thus, a change in eligibility

requirements will not upset the substantive state policies embodied in the domestic relations

law.

Indeed, in other contexts, the devolution of eligibility determinations is often celebrated

See Clayton P. Gillette, The Exercise of Trumps by Decentralized Governments, 83 VA. L. REV. 1347 (1997).

This is not to say that current eligibility requirements do not embody substantive policy preferences, only that once eligibility

requirements are established, they will rarely have significant effects on substantive domestic relations law. Of course, there are

some substantive state policies that are implicated by eligibility requirements, for example, the requirement that only unrelated

persons (of a certain level of consanguinity) are permitted to marry, or the limitation of marriage to two adult persons, or the

Hosted by The Berkeley Electronic Press

on the grounds that local administrators are better positioned to determine how government

benefits provided by higher-level governments should be distributed.

This bifurcation of tasks

takes advantage of both the local jurisdiction’s interest and expertise in choosing beneficiaries

and the state or federal governments’ interests in uniformity and economies of scale.

To be sure, local marriage eligibility rules may increase or decrease the number of

couples who can access a whole range of state-level benefits. For example, San Francisco’s

recognition of same-sex marriage will increase by some factor the number of couples eligible to

receive favorable tax treatment on their state tax returns. But the state-wide budgetary or fiscal

costs of any particular marriage eligibility regime are highly indirect. Local decisions

concerning property tax rates or spending on local schools have much more significant state-

wide fiscal affects. Moreover, the argument that the state has a fiscal or economic interest in

defining marriage is somewhat specious; the state does not regulate marriage eligibility out of

concerns for the state budget.

The specific inter-jurisdictional effects of local eligibility determinations are also difficult

to identify or quantify. One can imagine border problems arising if neighboring localities were

forced to recognize eligibility requirements with which they do not agree , but if cross-border

limitation of marriage to those above a certain age. These requirements are often justified on grounds of health and safety or

protection of minors. The limitation on the number of parties to a marriage is likely integral to the domestic relations law more

generally, so a change in that criteria would likely have more significant costs. See infra pp. ?-?.

It is not unusual for lower-level governments, including local governments, to be charged with deciding who is eligible to

access government benefits made available (and ultimately administered) by higher-level governments. For example, the Section

8 Housing program devolves significant responsibility to local public housing authorities. See 24 C.F.R. § 5.100; 24 C.F.R. §

960.202. Administration of welfare benefits through the Temporary Assistance to Needy Families (TANF) block-grants has been

devolved from the federal government to the states and, in some cases, by states to local governments. See Matthew Diller, The

Revolution in Welfare Administration: Rules, Discretion, and Entrepreneurial Government, 75 N.Y.U. L. Rev. 1121, 1179-81

(2000) (describing devolution of TANF administration from states to local governments in California, Colorado, North Carolina,

Ohio, and Wisconsin). The National School Lunch Program leaves student eligibility determinations in the hands of local school

food authorities. 7 C.F.R. § 210.7(c)(1).

Cf. Goodridge v. Dep’t of Pub. Health, 798 N.E.2d 941, 964 (dismissing the state’s claim that restricting marriage to

heterosexuals has the goal of conserving scarce private and public resources). Indeed, according to Elizabeth Scott, marriage

should be encouraged because it relieves government of fiscal burdens: “Under a well-structured marital regime, government

benefits and protections serve as a quid pro quo for the couple's agreement to alleviate society's dependency burden.” Elizabeth

Scott, Marriage, Cohabitation and Collective Responsibility for Dependency, 2004 CHI. LEGAL FORUM 225, 252-53.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

recognition is not a problem (or a problem that can be solved

), then those effects disappear.

Numerous commentators have engaged in the debate about whether the Constitution requires

states to recognize marriages that do not comply with in-state public policy; the question is still

unsettled.

Assuming, however, that there is no constitutional requirement of cross-border

recognition and that neighboring localities and states can decline to enforce out-of-locality or

out-of-state unions that they find uncongenial, there will be limited costs associated with

localities imposing their preferences on outsiders.

Indeed, localities engage in numerous activities that impose much more direct costs on

neighboring jurisdictions. For example, I have been highly critical (as have others) of how the

local zoning power has been used by suburban communities to entrench regional economic

inequality by imposing costs on neighboring urban centers.

Many commentators have also

argued that suburban zoning regimes have contributed to urban sprawl with its attendant

environmental consequences. Local marriage eligibility determinations do not raised these kinds

of concerns; the cross-border costs of marriage eligibility are negligible or even non-existent

when compared with other powers that local governments routinely exercise.

Congress adopted the Defense of Marriage Act (DOMA) in response to concerns that states would be required to recognize

out-of-state same-sex marriages. Pub. L. 104-199 (codified as amended in 1 U.S.C. § 7 and 28 U.S.C. §1738C). Some argue that

DOMA is unconstitutional. See L

AURENCE H. TRIBE, AMERICAN CONSTITUTIONAL LAW 1247 n.49.

Though some scholars have claimed that the Privileges and Immunities Clause and/or the Full Faith and Credit Clause could be

used to force unwilling states to recognize out-of-state marriages, see, e.g., Larry Kramer, Same-Sex Marriage, Conflict of Laws,

and the Unconstitutional Public Policy Exception, 106 Y

ALE L.J. 1965 (1997); Mark Strasser, The Privileges of National

Citizenship: On Saenz, Same-Sex Couples, and the Right to Travel, 52 RUTGERS L. REV. 553 (2000), many commentators argue

that neither of the clauses requires states to recognize marriages that do not comport with in-state public policy. See Yoo &

Vulchev, supra note 19, and Lea Brilmayer, Full Faith and Credit, W

ALL ST. J., March 9, 2004. See also Patrick J. Borchers,

The Essential Irrelevance of the Full Faith and Credit Clause to the Same-Sex Marriage Debate, 38 C

REIGHTON L. REV. 353,

363. Though states usually apply the “place of celebration” rule in determining the validity of a marriage, it is also hornbook law

that a state “has the power to determine who may assume or occupy the matrimonial relationship within its borders, and, in this

regard, a state legislature is competent to declare what marriages will be void in its own state, notwithstanding their validity in

the state where celebrated.” 52 AM. JUR. 2D Marriage § 62 (2004). Ralph Witten, Full Faith and Credit for Dummies, 38

CREIGHTON L. REV. 465, 486 (2005) (stating that “there is also a consensus that the Full Faith and Credit Clause as currently

interpreted does not require states to give effect to same-sex marriages performed in other states”). The Creighton Law Review

recently sponsored a symposium that debated these issues. See Symposium on the Implications of Lawrence and Goodridge for

the Recognition of Same-sex Marriages and the Validity of DOMA, 38 C

REIGHTON L. REV. 233 (2005).

See, e.g., Schragger, The Limits of Localism, 100 Mich. L. Rev. 371 (2001); Schragger, Consuming Government, 101 Mich. L.

Rev. 1824 (2003).

Hosted by The Berkeley Electronic Press

Of course, in a regime where marriage status can change with residency, cross-border

institutions and businesses would have to be aware that residency changes might cause marital

status changes. Third parties must already comply with both residency-based and status-based

changes, however. For example, a business operating in a city that extends anti-discrimination

protections to gays and lesbians has different responsibilities in that jurisdiction than in a

neighboring jurisdiction without such laws. Divorce, remarriage, bearing children in and out of

marriage—all produce status changes that affect third parties’ rights and responsibilities. These

kinds of status changes already occur frequently, sometimes multiple times for one individual .

Thus, while the costs to third parties of complying with local status changes are not zero, they

are not demonstrably higher than other cross-border kinds of costs that a diversity of regulatory

regimes often impose.

A different kind of cost of local eligibility regimes might be what commentators have

labeled “intangible externalities”

—the cost imposed on those who do not share a local norm

that they find immoral. Often the reason given for state regulation of marriage eligibility is that

states have an interest in the perpetuation of the traditional family unit and a stake in affirming or

denying the morality of particular kinds of relationships. To the extent that San Francisco’s

recognition of same-sex unions creates a morally offensive legal regime, that is a concern for all

the citizens of California.

Whether the government should be involved in making assertions of collective moral

values that do not implicate tangible individualized harms is itself subject to debate.

But even

assuming that assertions of public morality are valid grounds for regulation, it is not clear that

Gillette, supra note 36, at 1397-1407.

It famously was the grounds for the debate between H.L.A. Hart and Sir Patrick Devlin. See PATRICK DEVLIN, THE

ENFORCEMENT OF MORALS (1965); H.L.A. HART, LAW, LIBERTY AND MORALITY (1963); For arguments that moral

disapproval without more cannot be grounds for discrimination, see Lawrence v. Texas, 539 U.S. 558 (2003) (O’Connor

concurring) and Goodridge v. Dept of Public Health, 798 NE2d at 973 (Mass. 2003).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

they require centralized, or state-level, regulation. Intangible externalities are generated by any

law that has moral content or that implicates moral value s—much of local authority to regulate

primary conduct would be subject to this objection, thus swallowing up decentralized

government.

To the extent that the morality argument rests on the mere assertion of cross-

border disagreement, it is self-defeating: in regulating centrally, the moralizers are imposing a

serious cost on those local communities that do not share their moral preferences. One needs a

better reason, aside from the fact that the recognition of same-sex marriage constitutes a moral

nuisance, to make a compelling argument for centralized regulation.

One could argue, in fact, that public assertions of moral values are more appropriately

made at the local level. Perhaps that is why the Supreme Court continues to apply a community

standards approach to obscenity law, why locals often determine whet her to adopt blue laws or

Sunday closing laws, and why enforcement of sin-related offenses often varies greatly from

jurisdiction to jurisdiction. Because there is significant division of opinion about the appropriate

content of public morality, government regulation that is based predominantly on moral instincts

seems more appropriately tasked to lower-level governments. The intuition that Congress should

not be in the business of adopting “morals legislation” flows from this relationship between

values and scale.

The real concern with local regulation of marriage eligibility does not seem to be any

specific cross-border affects, but rather a generalized fear of outliers, those jurisdictions that

might adopt “odd” or “subversive” rules for marriage eligibility. Same-sex marriage fits in this

See Gillette, supra note 37, at 1402.

This is, of course, a controversial argument. For a discussion of the legitimacy of justifications based in “public morality” as

opposed to justifications based in “public welfare,” see Peter M. Cicchino, Reason and the Rule of Law: Should Bare Assertions

of Public Morality Qualify as Legitimate Government Interests for the Purposes of Equal Protection Review, 87 G

EO L.J. 139

(1998). See also Lawrence v. Texas, 539 U.S. 558, 599 (Scalia, J. dissenting); Romer v. Evans, 517 U.S. 620, 644-45 (Scalia, J.

dissenting). My argument here is not that assertions of “public morality” are always inadequate to justify government regulation,

only that such assertions are inadequate standing alone to justify centralized regulation.

Hosted by The Berkeley Electronic Press

category for many people, which is why there is support for a federal constitutional amendment

to ban it. One can easily think of other “outlier” eligibility regimes that would be “offensive” to

large majorities, such as underage marriage or polygamy.

But the problem of outliers seems overblown. Our federal system has seen a remarkable

convergence of state laws regulating marriage eligibility—until the last few years, marriage

eligibility criteria were relatively uniform across the nation. The same can safely be predicted

for a regime of local regulation. While a few jurisdictions might create significantly different

rules for entering into marriage, most will not.

Indeed, competitive models of local

government predict that local regulatory regimes will reflect majority preferences, sorting

individuals into jurisdictions that best reflect their desires. If that is so, than we would not expect

or want localities to adopt similar eligibility practices unless preferences are highly

homogeneous throughout the country. Moreover, “experimental” practices in one jurisdiction

often influence actions taken in other jurisdictions. The “highly offensive” eligibility regimes

adopted by some “outlier” local communities might be the very regimes that a small group

believes are the most just and that the wider polity might embrace if given the time.

C. Internal Oppression

These considerations point away from centralized regulation as a means of protecting

outsiders from being injured by local regulations. They do not, however, address the concern

that local regulation will hurt insiders who cannot protect themselves. The distribution of

powers up the scale of government is often a response to concerns about local majoritarian

oppression. Justice Scalia, for example, has invoked Federalist 10 in warning about the threats to

rights-holders when local legislative processes are captured by vocal minorities or sympathetic

Given the power, it is likely that more localities would tighten-up the eligibility requirements than those that would loosen

them. One can imagine localities adopting waiting periods or counseling requirements for those wishing to obtain licenses in the

jurisdiction.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

entrenched majorities.

If one believes that local majorities might be more inclined than state

majorities to adopt marriage eligibility requirements that are oppressive to local minorities, then

one might argue for keeping eligibility rules at the state level.

It is far from clear that a state monopoly of marriage eligibility rules will better protect

local political, cultural, or racial minorities, however. Anti-miscegenation statutes, for example,

were adopted at the state level. And the current debates over same-sex marriage pit locals

seeking to extend benefits to a historically disadvantaged minority against hostile state-wide

majorities. In the case of same-sex marriage, centralization is more likely to injure homosexuals

as a group than to help them—witness the federal move to shore up heterosexual marriage

through the Defense of Marriage Act and the adoption at the state level of same-sex marriage

bans.

Indeed, the whole notion of local oppression itself may be overstated, at least in the

context of marriage eligibility. Scalia’s (and Madison’s) denigration of local political processes

and corresponding faith in larger-scale political processes has come under significant criticism.

Some have argued, in fact, that local political processes are less susceptible to capture by special

interests, either majoritarian or minoritarian, and that the easy ability for individuals to exit local

jurisdictions makes it unlikely that local majorities can get away with oppression for long.

See THE FEDERALIST NO. 10 (James Madison); Romer v. Evans, 517 U.S. 620, 645-47 (1996) (Scalia, J., dissenting); Lawrence

v. Texas, 539 U.S. 558, 602-03 (2003) (Scalia, J., dissenting); Gillette, supra note 37, at 1402-1407.

For a similar argument, see Stephen Clark, Progressive Federalism? A Gay Liberationist Perspective, 66 ALB. L. REV. 719

(2003); see also Lynn A. Baker, Should Liberals Fear Federalism?, 70 U. C

IN. L. REV. 433 (2002). To the extent that one

believes that bans on polygamy protect women and children, one might point to the fact that the Mormons were required to reject

polygamy in order for Utah to be eligible for statehood as an instance where the central government—in that case Congress—

imposed a rule that protected a vulnerable minority. I do not mean to argue that centralized regulation is never protective of

vulnerable minorities, only that in the context of marriage eligibility, it is not at all clear why one would favor centralization over

decentralization as an abstract matter.

See Gillette, supra note 37, at 1402-07; Vicki Been, “Exit” as a Constraint on Land Use Exactions: Rethinking the

Unconstitutional Conditions Doctrine, 91 COLUM. L. REV. 473 (1991). See generally, ALBERT O. HIRSCHMAN, EXIT, VOICE AND

LOYALTY: RESPONSES TO DECLINES IN FIRMS, ORGANIZATIONS, AND STATES (1970).

Hosted by The Berkeley Electronic Press

Local majorities can and do exclude or oppress un- or under-represented individuals or

groups.

But even if one is suspicious of local majorities, it seems strange to view the local

extension of marriage benefits to an otherwise traditionally disfavored class as an incidence of

local oppression. This is not to say that those San Francisco residents who are offended by

same-sex marriage do not experience harms. In fact, they may feel deeply alienated from their

community or from the institution of marriage. These costs are not trivial, but they are not of

constitutional magnitude: the extension of marriage eligibility to same-sex couples does not

invade otherwise protected rights or exclude some in the community from benefits that are

enjoyed by others. Indeed, local eligibility determinations can offer more protection for

politically unpopular sub-groups who are unable to muster majorities at the state level.

No one

is being particularly oppressed by the adjustment of marriage to include more couples in San

Francisco.

Children, however, often figure prominently in the discourse about same-sex marriage,

and would fit into the category of an unrepresented group that might need protection from local

majorities who would do them harm. Those who argue that children will be hurt by allowing

same-sex marriage are usually making arguments that same-sex couples cannot be good parents,

that children of same-sex couples will be psychologically damaged, or that there is something

morally injurious about not having a “mommy” and a “daddy.”

These kinds of arguments are controversial.

But whatever their substantive merit, they

give us little guidance concerning the appropriate level of government at which to regulate

See, e.g., Schragger, Limits of Localism

See Brian Bix, State Interest and Marriage – The Theoretical Perspective, 32 Hofstra L. Rev. 93, 107 (2003) (observing that,

compared to state regulation, local regulation lowers the cost to third parties who are excluded or offended by government

regulation due to the reduced cost of exit to a more favorable jurisdiction).

See Goodridge, 798 N.E.2d at 962-64 (rejecting the argument that restricting marriage to opposite-sex couples furthers

Massachusetts’s policy of ensuring that children are raised in the “optimal” setting).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

marriage eligibility.

The “harm-to-children” argument for centralized regulation assumes that

local majorities would seek to harm children either for the sake of doing them harm or because

they are mistaken, and not for the reasons that we often associate with majoritarian oppression—

to extract rents or to monopolize the distribution of goods or power in society. And the argument

also assumes that centralized regulators – particularly state legislators – will be more attentive to

the welfare of local children than will local legislators because children are better represented at

the state level rather than at the local level. While the defense of children who cannot be

protected by local actors is a relevant reason for the state to act, it is not at all clear why we

would expect states to be more responsive to the effects of marriage eligibility regimes on

children than would localities. Indeed, we might predict just the opposite.

A more persuasive argument for centralized regulation of marital eligibility is that

children need stable family lives and the existence of non-uniform, local eligibility regimes will

cause them harm in circumstances where repeated changes in residence result in repeated

changes of parental status. This argument urges uniformity, and like all uniformity arguments, it

cannot be stopped at state borders — its logical outcome is a national rule. Moreover, the

stability argument assumes that the limited recognition of a child’s parents as married in one

jurisdiction even if only for the time that the parents reside there is worse than no recognition at

all at any time in any jurisdiction. In the context of same-sex marriage, the child stability

argument cannot coexist with the child oppression argument; stability more naturally leads to the

state-wide recognition of same-sex unions rather than the state-wide denial of such recognition.

This is not to say that the state does not have an interest in protecting children, say through an eligibility rule that prohibits

minors from marrying. Age of consent is an area where the state’s interests in public health and welfare seem significantly

stronger than in the same-sex marriage context.

For instance, municipalities played a significant role in early child and public welfare regulation. See Howard E. Jensen, The

County as an Administrative Unit in Social Work, 2 J. of Soc. Forces 552, 555 (1924) (noting that Kansas City established the

first Board of Public Welfare in 1910 and fifty other cities followed in by the mid-1920s). Note also that local zoning is often

defended as a means of protecting children. See Village of Belle Terras v. Borras, 416 U.S. 1, 9 (1974); Village of Euclid v.

Ambler Realty Co., 272 U.S. 365, 391, 394 (1926).

Hosted by The Berkeley Electronic Press

The very idea of decentralizing regulatory authority is to permit local majorities to enact

preferences that broader or more diffuse majorities might reject. Decentralizing the regulation of

marriage eligibility does not appear to put any unrepresented minorities at risk. In fact, it allows

a group that normally falls within that category – gays and lesbians — to obtain a more favorable

rule in a smaller jurisdiction.

In this way, local regulation of marriage eligibility is consistent with competitive models

of local government regulation.

If it is true that localities compete for a mobile citizenry, then

it makes sense to permit localities to differentiate themselves by regulating in areas that are of

true public concern. Marriage eligibility is a strong signal of the values of a particular

community; there is no reason that localities should not be able to compete for citizens on that

basis. The advantage, if the competitive model is correct, is that more individuals will find

governments that fulfill their preferences.

Participatory models of local government also support a local regulatory role. First, local

determinations of marriage eligibility are responsive to the values of community self-definition

and association; smaller-scale institutions are better suited to expressing the kinds of communal

norms implicated by marriage eligibility determinations. Second, marriage eligibility

determinations are high-profile governmental acts; moving the debate down a level of

government is a means of encouraging local self-government and civic engagement. To the

extent that citizens tend to ignore local government because of the perceived inconsequentiality

See, e.g., Charles M. Tiebout, A Pure Theory of Local Expenditure, 64 J. POL. ECON. 416 (1956). I do not wholeheartedly

endorse the competitive model, which, when taken to an extreme, results in significant inter-local economic inequality, see,

Richard C. Schragger, Consuming Government, 101 Mich. L. Rev. 1824 (2003) (reviewing W

ILLIAM A. FISCHEL, THE

HOMEVOTER HYPOTHESIS: HOW HOME VALUES INFLUENCE LOCAL GOVERNMENT TAXATION, SCHOOL FINANCE, AND LAND USE

POLICIES)

Professor Lynn Baker has made a version of this argument a number of times in advocating state-level diversity. See Lynn A.

Baker, Getting off the Dole: They the Court Should Abandon its Spending Doctrine, and How a Too-Clever Congress Could

Provoke it to Do So, 78 I

ND. L.J. 459, 474- 75 (2003); Baker, supra note 53, at 438-39.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

of its actions, decentralizing power to this sphere would enliven local debate and more directly

involve citizens in real issues with real consequences.

Indeed, in cases where there are no significant, state-wide or cross-border spillovers and

little fear of majoritarian oppression, then the benefits of inter-local competition and local self-

government seem to outweigh the state’s interest in asserting the value of a particular form of

intimate relationship through its eligibility rules. State-level regulation, while generating

uniformity, imposes significant costs on those individuals and communities who do not share the

majoritarian norm. When the statewide effects of devolving eligibility determinations to more

local institutions are relatively low, those exclusionary costs seem indefensible.

Of course, the argument for more diversity is not going to satisfy those who cannot

tolerate any diversity of marriage regimes, either because they believe that any departure from

the current eligibility requirements is immoral and destructive, or because they believe that

maintaining the current eligibility requirements is immoral and unconstitutional. The goal of a

national standard for marriage eligibility, either through constitutional adjudication or federal

constitutional amendment, is shared by both sides in the same-sex marriage debate. The

argument for local discretion does not address those kinds of claims directly—it merely suggests

that as a functional matter, local determinations may be superior to more centralized ones. It

seems that in this case smaller may be better.

II.

Constitutional Home Rule and Same-Sex Marriage

The previous section sought to persuade skeptics that marriage eligibility could and

maybe should be regulated as a matter of local concern. This section takes that argument

See Gerald A. Frug, The City as a Legal Concept, 93 HARV. L. REV. 1057, 1065-66 (1980).

See Bix, supra note 52.

Hosted by The Berkeley Electronic Press

significantly further by claiming that the Constitution might in fact require that marriage

eligibility determinations be left to localities, at least with regard to same-sex unions.

This argument is based on a localist reading of Romer v. Evans.

Recall that Romer

involved a Colorado constitutional amendment that barred the state, its agencies, and local

governments from adopting anti-discrimination ordinances that protected gays and lesbians.

Amendment 2 was intended to override then-existent state regulations and local ordinances or

policies that barred discrimination based on sexual orientation. In striking down the Amendment

as a violation of the Equal Protection Clause, the Court explicitly avoided holding that

homosexuals constituted a suspect class. Instead, it held that the Amendment failed to comport

with equal protection because it had no rational basis.

I understand Romer as being less about the individual rights of gays and lesbians not to

be discriminated against than it is about the collective right of the citizens of a local government

to decide, despite contrary state preferences, to protect gays and lesbians from discrimination.

Romer constrains states by providing for some sphere of “constitutional home rule”

for local

governments where the exercise of local authority is necessary to vindicate constitutional rights.

The argument that cities must have the discretion to adopt marriage eligibility criteria that

include same-sex couples follows from this Romer-based grant of constitutional home rule.

A. The Localist Reading Of Romer

The localist reading of Romer asserts that the Court’s problem with Amendment 2 was a

function of Colorado’s choosing an improper level of government to enact its ban on anti-

discrimination legislation protecting gays and lesbians. Colorado’s constitutional amendment

failed rational basis scrutiny because it prevented local gay-friendly majorities from adopting

517 U.S. 620 (1996).

See, e.g., Barron, supra note 22; Rosenthal, supra note 22.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

anti-discrimination legislation. The Court’s holding in Romer can thus be read as fairly narrow.

The Court did not compel the state, its agencies, or local governments to protect gays and

lesbians from discrimination nor did the Court hold that in every case legislation that

affirmatively prevented such ordinances would be ruled invalid. Instead it held that the state

could not prevent gays and lesbians from receiving anti-discrimination protections in those local

communities where they could muster majorities in support of such protections.

This localist account of Romer helps to explain the case’s particular application of

rational basis review. The Court was undoubtedly uncomfortable with Amendment 2 because it

“withdraws from homosexuals, but no others, specific legal protections from the injuries caused

by discrimination, and it forbids reinstatement of these laws and policies.”

But what made the

enactment illegitimate was that it both applied only to a small, traditionally disfavored group and

did so with an ill-fitting piece of legislation. The adoption (and constitutional entrenchment) of a

no-protective legislation rule at the state level constituted a violation of equal protection because

it was both too narrow and too broad: it targeted a particular group for a special political

disability by entrenching a statewide rule that preempted local pro-gay majorities. On this

account, Amendment 2 constituted a special disability because homosexuals and only

homosexuals had to resort to state political processes to seek the same kinds of protections

afforded to other non-suspect groups, “no matter how local or discrete the harm, no matter how

public and widespread the injury.”

This special “political” or, as Justice Scalia put it in dissent,

“electoral-procedural” discrimination,

had no rational basis: the “sheer breadth” of the

Romer, 517 U.S. at 631.

Id.

Id. at 639 (Scalia, J., dissenting).

Hosted by The Berkeley Electronic Press

disability, asserted the Court, is “so discontinuous with the reasons offered for it that the

amendment seems inexplicable by anything but animus toward the class it affects.”

Scalia, of course, invoked “electoral-procedural” discrimination as a way of dismissing

the Romer majority’s application of rational basis as both novel and illegitimate. And certainly

many commentators have found Romer’s reasoning to be so opaque as to raise the implication

that the Court’s opinion rests upon nothing more than a distaste for overt discrimination against

homosexuals.

The localist reading of Romer, however, may be the best way to understand how

equal protection can be offended on rational basis review in the absence of a suspect class and

without direct evidence of group-based animus.

In fact, the Sixth Circuit adopted this reading of Romer in Equality Foundation of

Greater Cincinnati v. City of Cincinnati.

Equality Foundation involved a city charter

amendment that was almost identical to the state constitutional amendment at issue in Romer.

The Sixth Circuit considered and upheld the charter amendment prior to the Romer decision; the

Supreme Court then ordered a rehearing in light of Romer, and again the Sixth Circuit upheld the

Charter amendment, this time distinguishing Romer on its facts.

The Equality Foundation court distinguished Colorado’s Amendment 2 from Cincinnati’s

charter amendment on the ground that the Colorado amendment was a state enactment while the

charter amendment was adopted by, and only applied to, the citizens of Cincinnati. As the court

Id. at 632.

Criticisms of Romer abound. See, e.g.,Lino A. Graglia, Romer v. Evans: The People Foiled Again by the Constitution, 68 U.

COLO. L. REV. 409 (1997); John C. Jefferies, Jr. & Daryl J. Levinson, 86 CALIF. L. REV. 1211, 1226-31; Louis Michael

Seidman, Romer’s Radicalism: The Unexpected Revival of Warren Court Activism, 1996 S

UP. CT. REV. 67 (1996).

128 F.3d 289 (1997), cert. denied 525 U.S. 943.

The Supreme Court subsequently denied the petition for a writ of certiorari. Justice Stevens wrote a one-page opinion joined

by Justices Souter and Ginsburg indicating some discomfort with that decision, however. Stevens wrote that “confusion over the

proper construction of the city charter counsels against granting the petition for certiorari,” and that “the Court's action today

should not be interpreted either as an independent construction of the charter or as an expression of its views about the underlying

issues that the parties have debated at length.” 119 S.Ct. 165, 525 US 943 (1998). See generally Mark Strasser, From Colorado

to Alaska by Way of Cincinnati: On Romer, Equality Foundation, and the Constitutionality of Referenda, 36 Hous. L. Rev. 1193

(1999).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

repeatedly stated, the Cincinnati charter amendment applied at the “lowest level of

government.”

This fact was crucially important because, according to the Sixth Circuit,

Romer held that Amendment 2 was illegitimate because it served to deprive “a politically

unpopular minority, but no others, of the political ability to obtain special legislation at every

level of state government, including within local jurisdictions having pro-gay rights

majorities.”

The state-wide adoption of a no-solicitude rule “was simply so obviously and

fundamentally inequitable, arbitrary, and oppressive.”

The same could not be said of an

equivalent local rule.

In contrast to Colorado, Cincinnati could generate pla usible reasons for adopting a local

charter amendment banning the city or its agencies from adopting anti-discrimination protections

for gays and lesbians. First, unlike the state, the locality had direct interests in deciding whether

it would expend local resources to protect gays and lesbians and whether there was sufficient

local objection to gay and lesbian accommodation . As the Sixth Circuit noted, the city had both

fiscal and associational interests that the state did not share.

And second, the charter adoption

did not impose any special procedural disabilities on gays and lesbians because they were not

forced to resort to statewide processes to seek redress. The Charter Amendment could be

repealed if opponents successfully convinced the citizens of Cincinnati to agree to do so (which

is in fact what happened in November 2004

). In contrast, Colorado Amendment 2 disabled

gays and lesbians and their allies from accessing the local political processes by forc ing them to

seek a state constitutional amendment even if they had the support of a local majority. While “a

state law which prevents local voters or their representatives, against their will, from granting

Equality Foundation, 128 F.3d at 297.

Id.

Id. at 300-01.

Kevin Osborne, Issue Results Surprise Pundits, CINCINNATI POST, Nov. 4, 2004 at A1.

Hosted by The Berkeley Electronic Press

special rights to gays, cannot be rationally justified by cost savings and associational liberties

which the majority of citizens in those communities do not want,”

the same cannot be said of a

local one.

The Sixth Circuit thus embraced the “electoral-procedural” reading of Romer. City

charter amendments barring anti-discrimination protection do not violate equal protection

because the locality has in a sense “internalized” the costs and benefits of the legislation

—both

those injured and those benefited by the restriction share the same political and territorial space.

Identical state constitutional amendments, in contrast, violate equal protection because the state

has not internalized the costs of the legislation, which will be borne only by gays and lesbians in

places where they would otherwise have received the benefits of local protection. The difference

between state and local legislation is captured in the court’s assessment of the asserted rationale

for the action. While the government may have a rational basis for denying anti-discrimination

protections to gays and lesbians, it does not have an interest in also putting them (or other

discrete social groups) at a particular political disadvantage. The substantive injury of the denial

of baseline rights in Romer was coupled with the procedural insult of political entrenchment.

Taken together, the two could only be indicative of animus, “a bare desire to harm homosexuals,

rather than to advance the individual and collective interests of the majority of Colorado's

citizens.”

B. The Contours of Constitutional Localism

Equality Foundation, 128 F.3d at 300.

I borrow this use of “internalization” from Kahan and Meares, who have made arguments about legislative costs and benefits

in the course of defending neighborhood-level anti-gang ordinances from constitutional attack. See Tracey L. Meares & Dan M.

Kahan, The Wages of Antiquated Procedural Thinking: A Critique of Chicago v. Morales, 1998 U. C

HI. LEGAL F. 197. While I

disagree with their defense, see Richard C. Schragger, The Limits of Localism, 100 M

ICH. L. REV. 371 (2001), I think the idea of

“internalization” is a useful one.

Equality Foundation, 128 F.3d at 299 (quoting Romer, 517 U.S. at 632).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

The local/state distinction is not fully theorized by the Romer Court or the Sixth Circuit;

Romer in particular, is a difficult opinion to parse. But the logic of a localist reading of the two

cases can be teased out. First, Romer tells us (if we adopt the Sixth Circuit’s view) that the

Equal Protection Clause operates differently at different levels of government. Identical

substantive enactments may have different constitutional effects depending on the source and site

of adoption. This means, second, that there may be circumstances under which the Constitution

requires that localities be free from state preemption. Because the Equal Protection Clause bars

the state from acting to override local decisions under certain instances, localities may enjoy a

form of constitutionally-mandated “home rule” that is incidental to the protection of

constitutional rights. After Romer, Boulder, Colorado can keep its anti-discrimination ordinance

despite the state’s objection.

This form of constitutional home rule is not foreign to the Court’s equal protection

jurisprudence. In Washington v. Seattle School District, the Supreme Court struck down a

Washington state constitutional initiative that prevented local school districts from adopting

voluntary desegregation plans involving intra-district busing.

The Court held that the state-

level initiative violated the Equal Protection Clause because it restructured the political process

by taking authority away from local school districts to remedy racial imbalances and by “lodging

decisionmaking authority over the question at a new and remote level of government.”

Though

Seattle School District con cerned the procedural burdens that a shift in decision-making

authority imposed on a suspect class, the Court’s reasoning parallels the localist logic of Romer-

Equality Foundation. Under Seattle School District, identical substantive enactments at the state

and local level are treated differently for equal protection purposes. Writing in dissent, Justice

458 U.S. 457 (1982).

Id. at 483.

Hosted by The Berkeley Electronic Press

Powell objected to this differential treatment, observing that while the state’s ban on voluntary

busing was constitutionally invalid, an identical local ban would not be.

Powell correctly

concluded that under the majority’s reasoning local governments were immunized from contrary

state commands even when their actions were not constitutionally compelled. Racial minorities,

objected Powell, thus appear to have a “vested constitutional right to local decisionmaking.”

Unlike Seattle School District, neither Romer nor Equality Foundation involve a suspect

classification. Their holdings therefore need to rely on a more robust account of local authority.

Both cases thus beg a number of questions. In states where local governments currently have no

power to adopt general anti-discrimination ordinances, does Romer require that localities receive

such powers? Or does Romer merely stand for the proposition that in states where localities have

the general authority to do so, the state cannot override that authority with reference to one

traditionally disadvantaged group? Does Romer-Equality Foundation imply a general rule of

state non-interference in local affairs, at least where equal protection values are implicated? Or

do the cases simply require that when states regulate, they do so through general legislation?

One possibility is that Romer has very little at all to say about the extent of constitutional

home rule protection. Conventional constitutional doctrine has always treated localities as

instrumentalities of their states, without independent constitutional status.

To the extent that

Romer departs from this background assumption, it only holds that in those states where local

governments are generally permitted to adopt anti-discrimination legislation, the state cannot

take away local authority to adopt such legislation for gays and lesbians. Romer thus offers

some protection to local governments but only when the state has already allowed the locality to

Id. at 494 (Powell, J., dissenting).

Id. at 489-99. see also Coalition for Economic Equality v. Wilson, 122 F.3d 692 (9th Cir. 1997) (adopting Powell’s dissenting

views in rejecting a challenge to California’s Proposition 209 which banned all governments in California from adopting

affirmative action programs for women and minorities).

See, e.g., Holt Civic Club v. City of Tuscaloosa, 439 U.S. 60, 71 (1978) (noting that state legislatures have “extraordinarily

wide latitude” in controlling the political reach of local governments).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

regulate. Thus, a state could prevent its cities from protecting gays and lesbians by preventing

them from adopting any anti-discrimination legislation whatsoever. This deference to state

power would be consistent with the rule in Equality Foundation, which takes Cincinnati’s power

to adopt charter limiting legislation for granted. Equality Foundation does not appear to limit the

state’s ability to override Cincinnati’s ban, it only holds that there are no federal constitutional

barriers to such an enactment if previously permitted by the state. Romer then would constitute a

limited restriction on how states organize and distribute their powers internally, but it would not

recognize local governments as independent political actors.

A more expansive reading of Romer-Equality Foundation is possible, however—one that

derives from Romer’s implicit protection of local political processes a nascent constitutional

basis for ensuring local political autonomy under certain circumstances. David Barron, for

instance, has argued that a constitutional norm of localism can be derived from the structure of

constitutional self-government.

Borrowing from Thomas Cooley’s nineteenth century

articulation of a right of local self-government, Barron argues that local political autonomy may

be necessary to vindicate substantive constitutional norms, in particular those “constitutional

rights [that are] partially dependent upon local political action.”

Barron thus reads Romer as a

step towards judicial recognition of localities as politically and constitutionally-salient

institutions. What is important about Romer is that it protects the ability of local governments to

extend equal treatment norms to a group that would not otherwise be entitled to those

protections. Local governments, on this account, would have a role in securing affirmative or

positive rights beyond those that would be protected through the normal course of constitutional

adjudication.

Barron, supra note 22.

Id. at 603.

Hosted by The Berkeley Electronic Press

Barron’s account implies that Romer grants local governments the power to adopt sexual

orientation anti-discrimination policies as a matter of federal constitutional law, and protects

them from contrary state commands. This rule follows from the Supremacy Clause: the

constitutional command of equal protection is enforced through a decentralized grant of authority

to local governments.

Although Barron resists the notion that one can define a constitutionally

protected sphere of local government sovereignty that exists “for its own sake,”

he does argue

that localism has a constitutional source: those underenforced constitutional norms that local

governments are uniquely positioned to protect. “[L]ocal . . . freedom from state law constraints

merits federal constitutional protection when such recognition would serve some independent

substantive constitutional value.”

Read broadly therefore, Romer would protect some realm of local control. To the extent

that states override local decisions that they do not like without reference to legitimate state

interests, they are violating a constitutional norm of local self-governance that is implicit in the

structure of constitutional rights. As a matter of federal constitutional law then, all localities

would have the power to adopt ordinances protecting gays and lesbians, regardless of state

constitutional provisions limiting their authority more generally.

C. State Bans on Same-Sex Marriage

The implications of either of these two readings of Romer – the first narrow, the second

broad — for those interested in the constitutionality of state-level same-sex marriage bans is

obvious. Those bans are structurally similar to Colorado’s Amendment 2: They constitute state

constitutional entrenchments that disable homosexuals from receiving the benefits of marriage

Id.; see also Roderick Hills, Dissecting the State: The Use of Federal Law to Free State and Local Officials from State

Legislatures’ Control, 97 M

ICH. L. REV. 1201 (1999).

Barron, supra note 22, at 602.

Id. at 600.

Lawrence Rosenthal has also argued for such a doctrine, though he offers a more schematic version that is based in an

expansive notion of equal protection. See Rosenthal, supra note X.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

unless homosexuals can convince a majority in the state to revisit that determination. Of course,

one might argue that there is an important threshold difference between an amendment that

“affirms” the heterosexual nature of marriage and an amendment that refuses to allow any

government unit to protect homosexuals from discrimination. This distinction seems

unconvincing, however. In light of the history of same-sex marriage bans, the claim that they do

not target gays and lesbians seems specious. And no one argues that the current marriage regime

does not discriminate against homosexuals. For present purposes therefore, I will assume that

two elements decisive to the decision in Romer are present in state same-sex marriage bans: a

traditionally disfavored class and state-level entrenchment.

Is there animus, however? One way to answer this question, according to Romer and

Equality Foundation, is to determine whether the state is acting to override local majorities

without good reason. On the limited version of localist constitutionalism described above, the

state can defend itself quite easily. Because marriage eligibility requirements have never been

locally determined, the state’s ban on same-sex marriage does not constitute a special political

disability imposed upon a historically disadvantaged group. Same-sex marriage bans do not

eliminate powers that local governments formerly possessed, and thus gays and lesbians are not

being deprived of a political venue that they otherwise would have had. In other words, the

political costs and benefits of the same-sex marriage ban are not being externalized: they are

imposed at the state level, as always.

Under the more robust version of localist constitutionalism described above, however, the

fact that localities have never been empowered to adopt marriage eligibility rules is not

determinative. Instead, one might look at the reasons why San Francisco embraced same-sex

marriage, and whether the city’s recognition of such marriages was an effort to promote

Hosted by The Berkeley Electronic Press

important substantive constitutional values. Certainly city officials, including the mayor, thought

that they were acting consistently with constitutional requirements. To the extent that the

recognition of same-sex marriage promotes equality by permitting a disfavored group access to

civil rights available to most others, one could argue that Romer requires a local option to

vindicate an underenforced constitutional norm of equal treatment.

But merely stating the case does not make it so. It turns out to be quite difficult to figure

out when local power is necessary to vindicate an important constitutional value. It is also

unclear why courts cannot simply vindicate that value directly. If constitutional equality is

offended by same-sex marriage bans, then courts should strike them down with no particular

reference to local governments.

The requirement of a local option, however, may turn on the difference between the

exercise of state and local power as a matter of substantive equal protection doctrine. A ban on

same-sex marriage at the state level might indicate a kind of animus that is not present in the

local context, as the court found in Equality Foundation. To figure out if this is so, one would

examine the state’s reasons for overriding the local extension of marriage benefits to same-sex

couples, or for regulating marriage eligibility at the state level at all. The validity of same-sex

marriage bans would thus turn on an independent assessment of the appropriate level of

government for regulating marriage eligibility. Under this version of Romer and Equality

Foundation, equal protection requires that local governments have the authority to issue same-

sex marriage licenses if the state cannot come up with good reasons why it needs to override the

local eligibility rules.

Animus, on this reading of Romer and Equality Foundation, is

evidenced by the failure of the state to offer good reasons for regulating centrally.

See, for example, Rosenthal’s equal protections analysis in id.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

The rule can be articulated as follows. Where a local regulatory choice that grants equal

benefits to a normally unpopular group—i.e., San Francisco’s recognition of same-sex

marriages—appears not to produce internal or external effects that require statewide regulation,

then courts should take a hard look at state rules that override that choice—i.e., same-sex

marriage bans. The combination of the absence of good reasons for centralized regulation, the

unpopularity of the group, and the group’s ability to obtain some measure of protection from

local majorities will be indicative of state-wide animus, an impermissible motive for government

regulation. As a consequence, the state regulation will be struck down, preserving a realm for

local discretion—in this case, the discretion to recognize or not to recognize same-sex marriages.

Of course, states can defend same-sex marriage bans on the grounds that they do have

good reasons to regulate centrally – that is, states can dispute my conclusion in Part I that marital

eligibility determinations are not properly made at the state level. But states have to make the

argument, whereas localities — which are also not free to adopt laws explainable solely by

animus — benefit from traditional rational basis review. That is because in the absence of a

suspect class, the heightened rational basis of Romer only kicks in when there is a traditionally

disfavored class and state-level regulation. State-wide regulations that disfavor homosexuals

will fail if the state acts unreasonably in overriding local efforts to extend benefits to them. And

in the context of same-sex marriage the question of reasonableness ultimately turns on the

appropriateness of a state rule that regulates marital status, the question I addressed in Part I.

This form of localist constitutionalism understands Romer-Equality Foundation’s

attention to the scale of government action as integral to the equal protection analysis. Recall that

the electoral-procedural reading of Romer—which the Sixth Circuit appeared to adopt in

Equality Foundation—requires that courts examine the level of government at which legislation

Hosted by The Berkeley Electronic Press

is adopted to determine if legislatures are isolating a particular group for special adverse

treatment. The concern that state legislatures may act too particularistically is not new: that fear

animated mid-19

century state constitutional bans on special legislation, which were meant to

prevent state legislatures from passing targeted laws that favored certain groups over others.

Romer’s electoral-procedural reading reflects the notion that laws should apply generally; equal

protection itself, as interpreted by the Romer court, can be understood as a limit on particularistic

enactments that seem not to be responsive to an identifiable functional state interest.

Nevertheless, in order to embrace a Romer-Equality Foundation-style constitutional

localism, one has to reject two pieces of conventional constitutional wisdom: first, that rights do

not vary across levels of government, and second, that states exercise plenary power over their

political subdivisions. The notion that localities might be differently situated than states in

respect to constitutional guarantees — that there may be constitutional space for local regulation

where states are not permitted to tread — challenges both those who express concern that rights

should not be differentially applied across levels of government as well as those who are

concerned about preserving state power.

III.

Should the Constitution Protect Localities From States?

Should the Court protect localities from contrary state commands when vindicating

particular substantive constitutional guarantees or a “right” of local self-government? One might

expect a Court committed to preserving the difference between “what is truly national and what

is truly local,”

to be sympathetic to local claims for constitutional or federal protection against

overbearing states. But the exact opposite is true: state plenary control of local governments is

Jeffrey M. Shaman, The Evolution of Equality in State Constitutional Law, 34 RUTGERS L. J. 1013, 1043-44 (2003); H.

Jefferson Powell, The Lawfulness of Romer v. Evans, 77 N. C. L. Rev. 241, n.52 (1998); Melissa L. Saunders, Equal Protection,

Class Legislation, and Color Blindness, 96 MICH. L. REV. 245, 251-62 (1997).

See United States v. Morrison, 529 U.S. 598, 617 (2000); United States v. Lopez, 514 U.S. 549, 567-68 (1995).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

the constitutional rule, not the exception, and a committed federalist will side with the states

against both federal intervention and local disagreement.

Localism is not an aspect of

federalism. When states and localities disagree, federalism and localism are conceptually

incompatible, and federalism, more often than not, wins.

This is notable because many of the justifications for federalism — democracy,

accountability, local control and experimentation — tend to support devolving power beyond the

states. To the extent smaller units can capture more of the benefits of self-government, localities

seem even better positioned to make the case for the exercise of decentralized power than are

states.

Nevertheless, those who advocate a doctrine of constitutional localism face a number of

objections. One objection is the concern that localities are too small to be trusted with

significant political decision-making power. This objection sounds in the Madisonian fear of the

politics of small places that I mentioned earlier – the notion that smaller polities will be more

homogeneous and thus more susceptible to majoritarian faction. Of course, many localities,

especially cities, are larger and more diverse than many states. But even if most localities are

generally smaller and less diverse, the claim that they are more susceptible to faction than states

is overstated. States have been the sites for the worst kinds of majoritarian oppression in

American history — slavery and Jim Crow – and public choice scholars have identified state and

federal governments as rife with special interest capture. Indeed, the dominance of factions in

our federal political system indicates that Madison may simply have been wrong about who wins

and loses in a politically enlarged sphere.

See Schragger, Reclaiming Canvassing Board, supra note 27 at 419-21.

Id. at 424.

See Frank H. Easterbrook, The State of Madison’s Vision of the State: A Public Choice Perspective, 107 HARV. L. REV. 1328,

1333-39 (1994); Cass R. Sunstein, Interest Groups in America Public Law, 38 S

TAN. L. REV. 29, 48-49 (1985).

Hosted by The Berkeley Electronic Press

Scholars have argued that local governments, in contrast, are constrained in their ability

to oppress minorities by the ability of those minorities to exit the jurisdiction.

More generally,

local governments may experience a special kind of economic and social pressure to avoid

regulatory extremes of any kind, including those that favor certain groups over others.

Anecdotal evidence at least indicates that local communities are often quite concerned about

their presentation to outsiders, and will seek to avoid taking controversial positions that might

put them at a disadvantage in the competition for local growth or economic development.

I do

not mean to argue that local governments are never afflicted with faction – they often are – but

only that from a comparative perspective, if one desires the broad distribution of political power,

local governments look at least as competent as states to exercise it.

A different objection to a judicial doctrine of constitutional localism is that the

Constitution already sufficiently protects local power by protecting state power. This “federalism

protects localism” argument is a variant of the political safeguards argument in federalism

debates—the assertion that state legislatures will protect locals’ desires for autonomy parallels

the claim that Congress will protect states’ desires for autonomy.

On this argument, the

Supreme Court should not defend locals against states because on balance a robust federalism

will preserve local power better than judicial forays into the state-local relationship.

Roderick Hills, for example, argues in his contribution to this symposium that a federal

system of government tends to do a better job at devolving power downward to localities than

See, e.g., Carol M. Rose, The Ancient Constitution vs. The Federalist Empire: Anti-Federalism from the Attack on

“Monarchism” to Modern Localism, 84 N

W. U. L. REV. 74, 100 (1989).

See, e.g., Jodi Wilgoren, Vote in Topeka Today Hangs on Gay Rights and a Vitriolic Local Protester, NY Times, March 1,

2005, at A12 (reporting that residents of Topeka are concerned about being perceived as either “a haven for homophobia or for

homosexuality” in voting for candidates for city council). Commentators have argued that the median voter model best captures

the politics of local government. See, e.g., W

ILLIAM A. FISCHEL, THE HOMEVOTER HYPOTHESIS: HOW HOME VALUES INFLUENCE

LOCAL GOVERNMENT TAXATION, SCHOOL FINANCE, AND LAND-USE POLICIES (2001)

For a sophisticated discussion of the political safeguards argument, see Larry D. Kramer, Putting the Politics Back into the

Political Safeguards of Federalism, 100 C

OLUM. L. REV. 215, 223-28 (2000).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

does a unitary system of government.

Hills argues that states will be less wary about

distributing power downward in part because states can more easily retrieve such powers (if they

so desire) than can a national government. But the paltry grants of power in even those states

with ample home rule provisions seem to indicate otherwise. States may, in fact, resist the

expansion of local authority much more so than would a national government because local

authority is more of a threat to state prerogatives than it is to national ones.

Whether localities do better in a federal system than in a unitary one is an important

question, but not exactly on point: those advocating localism are not seeking to replace our

current federalism with something akin to a European unitary system. The most robust version

of constitutional localism co-exists with federalism: the constitutional right of local self-

government is cashed-out as a protected sphere of authority for sub-state political units, defended

against both state and federal encroachment.

This brings me to a final objection, which I believe is the most serious. The primary

problem with a robust version of constitutional localism is not that localities cannot be trusted or

that states already do an adequate job in protecting them. Rather, constitutional localism suffers

from a vulnerability that afflicts most attempts to create spheres of political authority: the

difficulty in arriving at a plausible or ultimately workable principle for allotting some powers to

one level of government and some to others. Federalism decisions allotting power between the

national government and states, as well as home rule decisions that apportion powers between a

state and its localities, both suffer from this defect. Courts have sought to articulate neutral

principles that can differentiate some exercises of power or arenas of expertise from others.

Roderick M. Hills, supra. For a contrary argument that asserts that local governments get less resources in a federal system,

see Pradeep Chhibber & E. Somanathan, Does Federalism Imply Devolution? Local Governments in Federal Nations (July 2002,

Draft).

For a discussion of the relationship between legal and political localism and an account of when local officials in non-federal

systems can be said to exercise significant local power, see E

DWARD PAGE, LOCALISM AND CENTRALISM IN EUROPE (1991).

Hosted by The Berkeley Electronic Press

These principles can rarely support the weight put upon them, however, and the results are often

arbitrary or simply foregone.

100

Commentators have long argued that the Court’s attempts to come to some workable

solution to this line-drawing problem have been abject failures.

101

Those failures should warn us

away from adding a third layer of government to the mix. Nevertheless, the solution to the line-

drawing problem is not to abandon constitutional localism, but rather to abandon it in its

proceduralist form. Instead of searching for some neutral principles by which to divide up

powers a priori, a robust localism can begin with the substantive rights at stake and figure out

how those rights are implicated when different levels of government regulate, as in Romer and

Equality Foundation.

This approach does not solve all the problems attendant to differentiating government

powers: the hard part in Romer-Equality Foundation is figuring out why the state’s adoption of

an anti-gay constitutional amendment violates the Equal Protection Clause but the city’s

adoption of an equivalent charter amendment does not. But all this means is that there is work to

be done in sorting out how the scale of government regulation differently affects substantive

constitutional rights. That work can proceed more successfully within a discourse that focuses

on rights rather than on powers, not because rights discourse is any more deterministic, but

because it is grounded in actual affects on individual claimants.

Mark Rosen, for instance, engages in this kind of work in his contribution to this

symposium, as he has elsewhere .

102

Rosen argues that the level at which government regulates

100

See Edward Rubin & Malcolm Feeley, Federalism: Some Notes on a National Neurosis, 41 UCLA L. Rev. 903 (1994).

101

See, e.g., Garcia v. San Antonio Metro. Transit Auth., 469 U.S. 528 (1985) (overruling as “unsound in principle and

unworkable in practice” the line drawn in National Legue of Cities v. Usery, 426 U.S. 833 (1976))

102

Mark D. Rosen, The Surprisingly Strong Case for Tailoring Constitutional Principles, 153 U. PA. L. REV. (forthcoming 2005)

[hereinafter “Tailoring Constitutional Principles”]; Mark D. Rosen, Establishment, Expressivism, and Federalism , 78 Chi.-Kent

L. Rev. 669 (2003) (arguing that, in Establishment Clause doctrine, it may be desirable to “size” constitutional limitations to the

level of government that is acting).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

can be taken into account by courts when they determine the existence of a constitutional

violation—that the doctrinal tools currently exist to “tailor” constitutional rights to the level of

government regulation.

103

This approach has been proposed by scholars in other contexts as

well, though not using the same terminology. For example, Robert Ellickson has argued that the

constitutional principle of one-person, one-vote should not apply to local governments, though it

should apply to the states.

104

I have argued, albeit for quite different reasons, that Establishment

Clause norms should apply differently at the local level than at the state or national level.

105

The

Sixth Circuit’s decision in Equality Foundation represents a version of this form of constitutional

localism: it treats the local-ness of the city charter amendment as central to the equal protection

analysis, thus enabling the court to differentiate it from an equivalent state-level enactment.

The site of government action will affect constitutional rights differently in different

substantive areas. I have argued that local governments should be free to adopt school voucher

programs that include pervasively sectarian institutions but that states and the national

government should be barred from doing so because centralized government regulation poses

significantly greater dangers to religious liberty than decentralized regulation. In equal

protection doctrine, the level of regulatory activity might be invoked to show animus, as in

Romer, or to rebut it, as in Equality Foundation. This form of localist constitutionalism does not

seek to define a sphere of local and state conduct and then place particular government tasks

within it. Rather, it integrates the question of scale into the Court’s substantive constitutional

determinations. Scale is a relevant consideration because it affects the substance of

constitutional rights.

103

Rosen, Tailoring Constitutional Principles, supra note 94.

104

Robert C. Ellickson, Cities and Homeowners Associations, 130 U. PA. L. REV. 1519 (1982).

105

Richard C. Schragger, The Role of the Local in the Doctrine and Discourse of Religious Liberty, 117 HARV. L. REV. 1810

(2004).

Hosted by The Berkeley Electronic Press

A constitutional doctrine attentive to scale serves other purposes as well, separate from its

effects on the substance of constitutional rights. First, attention to constitutional scale vindicates

norms of democratic self-government. The recognition that local governments may be

differently situated to constitutional norms than state governments treats local political

communities seriously as sites for democratic participation. Both Romer’s preservation of a

local option to protect gays and lesbians and Equality Foundation’s preservation of a local option

to refuse broadly such protection can be understood as encouraging participatory democracy in

small-scale settings. If one’s goal is to develop deliberative democracy,

106

one way to do so is to

create political space for citizens in face-to-face communities to make significant moral and

political decisions.

Second, and relatedly, a doctrine of localist constitutionalism gives the Court a way to

express concern with the way state governments are treating certain of its members without

adopting a national rule that pre-empts all democratic decision-makers. The presence of a

traditionally disfavored class that would otherwise gain protections in a small number of

localities where they have the support of a majority indicates a potential political failure at the

state level. A localist constitutionalism is a mechanism for regulating these state political

processes.

Consider Romer then as offering the Court a choice between two possible solutions to the

problem of faulty state political processes. The first solution is to protect the group subject to

faulty state processes by centralizing the rules governing the group’s treatment, i.e., through a

constitutional rule. This approach is advocated by those who urge the Court to treat

homosexuals as a suspect class. The second solution is to protect the group subject to faulty state

processes by decentralizing the rules governing the group’s treatment, i.e., by giving the group a

106

See Stephen Breyer, Our Democratic Constitution, 77 N.Y.U. L. REV. 245 (2002).

http://law.bepress.com/uvalwps/uva_publiclaw/art30

real opportunity to gain protection through the local political process. The Court chose the

second solution and, in doing so, opened up a potentially fruitful new avenue for remedying state

level political failures by showing that solutions to such failures can come from above or from

below.

One would, of course, have to trace out the circumstances in which decentralization is a

better solution than centralization to political process failures in the states. One possibility is that

a localist equal protection doctrine acts as a break on state legislatures’ tendencies to favor

powerful, particularistic interests. Because laws adopted at higher levels of government might

be more susceptible to minority or majority faction, are more difficult to change and are not

easily avoided through exit, those laws must be written in general terms or have easily

ascertainable public purposes. State-wide legislation that appears to target vulnerable groups has

to meet a higher level of rationality. That is not because we trust local governments more than

we do states or the federal government, but because equal protection may depend importantly on

how easily a particular group can access the political process, to what degree a political majority

is imposing preferences on itself or on outsiders, and to what extent state law limits a diversity of

regulatory schemes that appear to be rights-enhancing.

This formulation of the relevant considerations for a decentralized equal protection

doctrine is, of course, contestable. My limited goal here is simply to put localities on the table as

viable alternative institutions for solving political failures in the states under certain

circumstances. In order to embrace this alternative, however, one has to recognize—as I have

been urging—that localities are not simply mini-states or instrumentalities of the states.

Localities are not perfect—they too suffer from process failures—but they may be differently

Hosted by The Berkeley Electronic Press

advantaged in relation to some groups and some rights than are the states and the national

government.

Indeed, a decentralized equal protection doctrine in which the level of government is a

relevant consideration can fill a space in constitutional adjudication where the traditional equal

protection classification scheme seems inappropriate, that is, where centralization appears

problematic as a solution to group oppression for political or conceptual reasons. Courts

applying constitutional guarantees often have to consider whether to adopt a constitutional norm

that applies across the entire country. Thus, in the dispute over same-sex marriage, one would

expect that courts will determine if the Equal Protection or Due Process Clauses of the

Constitution require that all states everywhere recognize same-sex unions if states are to have

any marriage law at all. Of course, as I have already noted, the Supreme Court might avoid

making that determination by holding that state constitutional amendments affirming that

marriage is restricted to one man and one woman do not target gays and lesbians for special

adverse treatment. Indeed, the majorities in both Romer and Lawrence insisted that marriage

was not on the block.

107

Romer, it could be argued, involved a non-solicitation rule specific to

gays and lesbians and the specific withdrawal of rights that had already been granted, and

Lawrence involved the criminalization of gay and lesbian conduct – an outright ban on private,

consensual acts. One can imagine that a politically pragmatic Court might adopt just such

readings of Romer and Lawrence because it is fearful of upending the states’ traditional marriage

rules at a time when most Americans seem to support them.

Applying a decentralized approach based in a localist reading of Romer to same-sex

marriage offers another alternative, however. First, such an approach vindicates a constitutional

norm of equal treatment that has already found voice in Romer and Lawrence, and that is

107

Romer, 517 U.S. at 634; Lawrence, 539 U.S. at 578.

http://law.bepress.com/uvalwps/uva_publiclaw/art30

consistent with the constitutional trajectory of equal protection. And second, a localist approach

to same-sex marriage does not require a rejection of traditional heterosexual marriage, but simply

permits local majorities to adopt inclusive legislation at their discretion.

This course of action might be politically and constitutionally more palatable than the

other two options: either rejecting outright homosexuals’ serious claims of unequal treatment or

dictating a nationwide norm that a significant number of Americans do not share. By adopting a

version of constitutional localism, the Court can express its discomfort with state-mandated

unequal treatment of homosexuals while simultaneously respecting the significant differences of

opinion regarding same-sex marriage by allowing such opinions to be expressed at the level of

community most appropriate to their assertion. Constitutional values of participatory democracy

and equal treatment would both be vindicated and the sting of countermajoritarianism would be

minimized. To the extent that one agrees that marriage eligibility for same-sex couples is

properly determined at the local level, a constitutional rule making it a requirement may be just

the right thing.

Conclusion

Courts and legislatures are hard at work making rules for marriage eligibility that apply

across large expanses of territory in many jurisdictions and for many people. Few have thought

about the appropriateness of the scale of these endeavors, however. Most have assumed that

state-level determinations are worthy of functional and constitutional respect. Both claims might

be wrong, however. First, as a functional matter, local determinations of marriage eligibility

may be superior to state-level determinations—they are at least decidedly not inferior. Second,

as a constitutional matter, local determinations of marriage eligibility might be necessary to

vindicate constitutional norms or solve political process failures in the states. A constitutional

Hosted by The Berkeley Electronic Press

rule requiring that states permit localities to regulate marriage eligibility so as to allow same-sex

couples to opt into the state’s marriage regime might be the best solution to the problem of state

oppression consistent with norms of democratic self-government and equal treatment.

Those on both sides of the debate who argue that same-sex marriage implicates universal

norms of morality or of equal treatment will not be satisfied with anything less than a national or

international rule. I am sympathetic to that argument. This Essay does not advocate short-

circuiting that debate, nor does it provide moral or constitutional reasons for siding with one side

or the other. I am not claiming that permitting local same-sex marriage opt-ins constitutes the

most just outcome for any individual claimant, only that localism is an important possibility in a

federal system that sometimes solves hard moral problems by decentralizing their solutions.

Cities like San Francisco and Boulder have begun to recognize themselves as constitutional

actors. Perhaps it is time for constitutional doctrine to do so as well.

http://law.bepress.com/uvalwps/uva_publiclaw/art30