39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         205 

WE ARE NOT STRUCK WITH BLINDNESS: THE 

ESTABLISHMENT CLAUSE AND RELIGIOUSLY 

MOTIVATED STATE PREEMPTION OF MUNICIPAL 

NON-DISCRIMINATION LAW 
 

KATHLEEN RITTER 
 

INTRODUCTION 

 

On August 19, 2019, the City Council of Fayetteville, Arkansas, had a ten-hour 

meeting.1 More than a hundred citizens lined up to share their thoughts on Ordinance 

119, a proposed regulation that had become a source of great public debate in the town. 

Only after they had all spoken did the meeting adjourn at 3:30 AM. The City Council 

then passed Ordinance 119 by a six-to-two margin, approving its establishment of 

protections on the basis of sexual orientation and gender identity in the city’s non-

discrimination law, and making Fayetteville the first town in the state to extend such 

protections to LGBT individuals.2 Six months later, the law was invalidated. The 

Arkansas General Assembly preempted Fayetteville’s legislation, making it outside of the 

city’s authority as a municipality to create protected classes beyond those that exist at the 

state level.3 Arkansas state law does not protect sexual orientation and gender identity. 

 

Preemption is a powerful tool that states can wield against municipalities, and in 

recent years they have been doing so with great frequency and occasionally great 

antagonism. The difficulty for municipalities is that preemption regimes often leave few 

channels through which cities can challenge preemptive laws and litigation has often 

relied on federal Constitutional claims when statutory and state constitutional claims are 

unavailable. Typically, these claims have been based on Equal Protection theories, 

especially where non-discrimination ordinances (NDOs) have been implicated. However, 

 
1 See Fayetteville Becomes First Ark. City to Pass LGBT Non-Discrimination Law, LGBTQ NATION (Aug. 

20, 2014), https://www.lgbtqnation.com/2014/08/fayetteville-becomes-first-ark-city-to-pass-lgbt-non-

discrimination-law/ [https://perma.cc/7P4Z-F6CB]. 

 
2 See Arkansas: LGBTQ Non-Discrimination in States, FREEDOM FOR ALL AMERICANS, 

https://www.freedomforallamericans.org/category/states/ar/ [https://perma.cc/7DJQ-T6WM]. 

 
3 See Rebecca Hersher, Arkansas Supreme Court Strikes Down Local Anti-Discrimination Law, NPR (Feb. 

23, 2017), https://www.npr.org/sections/thetwo-way/2017/02/23/516702975/arkansas-supreme-court-strikes-

down-local-anti-discrimination-law [https://perma.cc/H3KZ-PJZT]. 



 

206 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

as Professor Richard Briffault et al. note in their issue brief on preemption for the 

American Constitution Society, “the contemporary landscape of increasing state 

hostility” creates “an urgency to think[] creatively about new legal arguments.”4 One of 

these arguments, this Article proposes, is based on the Constitution’s Establishment 

Clause. The path to prevailing on Establishment Clause challenges to preemption law is 

steep but navigable, especially if courts are willing to adjust their jurisprudence with 

regard to both the Establishment Clause and preemption doctrine in general. 

 

As national politics become increasingly polarized and the federal government 

increasingly paralyzed,5 much of the work to protect the civil and political rights of 

citizens is happening at the municipal level. Municipalities, especially blue cities in red 

states, have in recent years been at the forefront of passing progressive legislation across 

many spheres, from minimum wage provisions and anti-discrimination ordinances to 

sharing economy regulations and gun laws. Though Justice Brandeis famously 

characterized states as “laboratories” of democracy, municipalities also serve as arenas 

for policy experimentation.6 Municipal governments are often more able to address local 

problems, respond directly to the concerns and preferences of citizens, and avoid the 

influence of spending by special interest groups than statewide governments.7 

Furthermore, policies adopted locally can catalyze policy changes at the state and 

 
4 RICHARD BRIFFAULT ET AL., AM. CONSTITUTION SOCIETY, THE TROUBLING TURN IN STATE PREEMPTION: THE 

ASSAULT ON PROGRESSIVE CITIES AND HOW CITIES CAN RESPOND (2017). 

 
5 The federal government was recently shut down for the longest period in its history, largely due to a 

partisan stalemate regarding immigration policy. See Saahil Desai, It’s Not Just the Longest Shutdown Ever, 

THE ATLANTIC (Jan. 13, 2019), https://www.theatlantic.com/politics/archive/2019/01/under-donald-trump-

longest-shutdown-ever/580114/ [https://perma.cc/RW79-GRSB]. 

 
6 New State Ice Co. v. Liebmann, 285 U.S. 262, 311 (1932). 

 
7 See Paul Diller, Intrastate Preemption, 87 B.U.L. REV. 1113, 1119–20 (2007). Diller notes that progressive 

groups more so than conservative ones tend to utilize “the smaller scale of city politics” in accomplishing 

their objectives rather than targeting their efforts at the statewide or national level. Id. at 1120. He identifies 

two explanations for this phenomenon: first, that the relative size of city politics as compared to national 

politics reduces the influence of campaign contributions by special interest groups, leaving municipalities less 

beholden to these interests and able to provide a “more responsive, democratic form of government” than 

state and national governments. Id. A second explanation for the proliferation of progressive policy 

experimentation at the local level may be that the largest cities in the United States, those most likely to have 

ambitious policy agendas, “have a strong leftist political tilt.” Id. (internal citations omitted). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         207 

national level.8 Policies that prove workable and popular on small, local scales may be 

adopted by other cities and eventually adopted at the state level.9 

 

This percolation of policy from cities “outwards” (to other cities) and “upwards” (to 

the state) is particularly observable in the civil rights sphere.10 Historically, localities 

have often led the charge in creating civil rights reform. For example, cities were at the 

forefront of the push for marriage equality. In 2004, San Francisco began issuing 

marriage licenses to same-sex couples, four years before marriage equality would become 

state law in California.11 San Francisco’s lead was promptly followed by Sandoval 

County, New Mexico; Multnomah County, Oregon; Asbury Park, New Jersey; and the 

mayors of both New Palz and Ithaca, New York.12 All of these decisions placed cities in 

direct conflict with state law, but highlighted the role that cities can play in moving 

forward state and national policy debates.13 Cities are also currently spearheading policy 

change in the areas of LGBTQ rights,14 climate change,15 public health,16 and 

immigration,17 among others. 

 
8 See id. at 1119. 

 
9 See Diller, supra note 7, at 1122. 

 
10 See id. at 1119. 

 
11 See Richard Schragger, Cities as Constitutional Actors: The Case of Same-Sex Marriage, 21 VA. J.L. & 

POL’Y 147, 148 (2005). 

 
12 See id. 

 
13 See id. for an in-depth analysis of the legal and constitutional role of municipal governments in the 

ongoing debate (at the time of its pre-Obergefell publication) concerning marriage equality. 

 
14 See, e.g., Lydia E. Lavelle, Grassroots Gay Rights: Legal Advocacy at the Local Level, 21 AM. U.J. 

GENDER SOC. POL’Y & L. 507 (2013); HUMAN RIGHTS CAMPAIGN & EQUALITY FED’N INST., MUNICIPAL 

EQUALITY INDEX: A NATIONWIDE EVALUATION OF MUNICIPAL LAW (2018) (assessing how inclusive municipal 

laws are of LGBTQ citizens). 

 
15 See, e.g., Oliver Milman et al., Four Cities Leading the Way Against Climate Change, CITYLAB (June 14, 

2019), https://www.citylab.com/equity/2017/06/the-fight-against-climate-change-four-cities-leading-the-way-

in-the-trump-era/530385/ [https://perma.cc/4QHD-46R9]. 

 
16 See, e.g., Abby Goodnough, In San Francisco, Opioid Addiction Treatment Offered in Streets, N.Y. TIMES 

(Aug. 18, 2018), https://www.nytimes.com/2018/08/18/health/san-francisco-opioid-addiction.html 

[https://perma.cc/6AKR-JQQJ]. 

 
17 See, e.g., Xi Huang & Cathy Yang Liu, Welcoming Cities: Immigration Policy at the Local Level, 54 URB. 

AFF. REV. 3 (2016); Janell Ross, 6 Big Things to Know About Sanctuary Cities, WASH. POST (July 8, 2015), 



 

208 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

When municipalities seek to fill gaps they see in statewide legislation, however, they 

run the risk of the state claiming that the local laws are preempted by state or federal law. 

Recently, states have begun using preemption doctrine rather aggressively, striking down 

local innovations by overriding municipal legislation or withdrawing authority from 

municipal governments.18 This phenomenon is particularly prevalent when cities seek to 

expand the boundaries of civil rights protections, though it is certainly not limited to 

conservative state government preemption of progressive municipal legislation.19 

 

Inevitably, tensions exist where, as in Arkansas, state legislatures and municipal 

governments have conflicting operative values; for example, when cities like Fayetteville 

seek to implement progressive legislation that conflicts with more conservative state 

governments, like that of Arkansas. In seeking to mount challenges to state action, cities 

and their defenders should be alert to the justifications provided by state legislatures for 

preemptive action. In most cases, state preemption of municipal action is valid and 

almost impossible to challenge, as will be discussed in Part II.B. On occasion, however, 

state preemption laws may run afoul of the federal Constitution. Preemptive legislation 

that is discriminatory, that unduly burdens vulnerable populations, or that is based on 

impermissible motivations—such as the establishment of religion—can and should be 

challenged. As Justice Hoffman of the Supreme Court of California observed in 1971 in 

Parr v. Municipal Court: 

 

When we take our seats on the bench, we are not struck with blindness, 

and forbidden to know as judges what we see as men; and where an 

ordinance, though general in its terms, only operates upon a special race, 

sect or class, it being universally understood that it is to be enforced only 

against that race, sect or class, we may just conclude that it was the 

intention of the body adopting it that it should only have such operation 

and treat it accordingly.20 

 

 
https://www.washingtonpost.com/news/the-fix/wp/2015/07/08/4-big-things-to-know-about-sanctuary-cities-

and-illegal-immigration/?utm_term=.85de25e25f3 [https://perma.cc/VA6S-VVBD]. 

 
18 See BRIFFAULT ET AL., supra note 4, at 1. 

 
19 See id. 

 
20 Parr v. Mun. Ct., 479 P.2d 353, 356 (1971). This case concerned a municipal ordinance prohibiting sitting 

on grass in public parks that was intended to discriminate against “hippies” and drive them out of the city. Id. 

at 355. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         209 

Justice Hoffman’s observation is as important now as it was in 1971. Throughout 

their history, courts in this country have been asked to distinguish between pretext and 

prejudice in the service of upholding the Constitution. In confronting legislation with an 

apparently religious motivation, even when that legislation claims to have a secular 

purpose, as most preemption laws do, courts should be able to exercise judgment in 

deciding not to defer to the purposes proffered by lawmakers. Arkansas’ preemption of 

Fayetteville’s NDO presents just such an opportunity to probe legislative motivation in 

the context of state-municipal preemption. 

 

This Article proceeds in three parts. Part I considers the doctrine of state preemption 

and the current landscape of state preemption of local power, focusing on the increase in 

state use of preemption in the last few years. Part II provides an overview of 

Establishment Clause jurisprudence, the state of the Supreme Court’s current practice 

regarding the Establishment Clause, and the elements of an Establishment Clause claim 

under the Court’s various tests. Part II also presents Arkansas’ Intrastate Commerce 

Improvement Act (“Act 137”) as a case study and suggests that, where evidence of 

impermissible religious legislative motivation exists, the Establishment Clause could 

provide an alternative to Equal Protection and other federal constitutional claims for 

challenging preemptive laws. Part III considers how well an Establishment Clause claim 

against Act 137 would fare, concluding that the probability of the bill being struck down 

on Establishment Clause grounds under current case law is low. As such, Part III 

proposes a revised Establishment Clause standard which borrows from Equal Protection 

case law. It suggests that, under this revised standard, an Establishment Clause claim 

against Act 137, and other bills with analogously suspect motivations, would have a 

much greater chance of being struck down. Part III also suggests that the current spate of 

aggressive preemption activity could potentially be mitigated if courts rethink their 

approach to state-local relations and preemption generally. 

 

I. Blue Cities, Red States: Preemption and Progressive Policies 

 

A. Recent Use of Preemption by States  

 

While preemption has long been a useful mechanism for states to advance legitimate 

state interests in uniformity and to push back against local parochialism, it has, of late, 

become a tool of conservative state government opposition to progressive local 

interests.21 According to a 2018 study published by the National League of Cities (NLC), 

 
21 See id. at 356; Diller, supra note 7, at 1114. 



 

210 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

state legislatures have “consistently . . . stricken down laws passed by city leaders in four 

crucial areas of local governance: economics, social policy, health and safety.”22  

 

The use of preemption by states to curtail local innovation has become rampant, with 

one scholar describing the amount of preempted legislation as having reached “epidemic 

proportions.”23 Both demographic and political trends indicate that many cities will 

continue to become more Democratic while their state governments remain in Republican 

control, suggesting that the use of offensive preemption to control progressive policy 

innovation at the municipal level will not abate any time soon.24 But this new preemption 

is not solely the product of demographic trends and the preponderance of Republican-

controlled state governments. Especially when red states seek to restrain blue cities, many 

preemptive bills can be traced to conservative think tanks and nonprofits, such as the 

American Legislative Exchange Council (ALEC) and the National Rifle Association 

(NRA), which not only lobby legislators to preempt progressive local action but also, in 

the case of ALEC, provide model legislation to do so.25 This increase in preemptive 

activity is concerning from the perspective of federalism and a concern for local 

autonomy, but it also has serious implications in areas such as civil rights, public health, 

and public safety.26 

 

 
22 NAT’L LEAGUE OF CITIES, CITY RIGHTS IN AN ERA OF PREEMPTION: A STATE-BY-STATE ANALYSIS 1 (2018). 

As of the most recent update, twenty-eight states have preempted minimum wage laws, twenty-three have 

preempted paid leave laws, forty-one have preempted laws regulating ride sharing, five have preempted laws 

regulating home sharing, twenty have preempted municipal broadband laws, forty-two have preempted tax 

and expenditure legislation, forty-three have preempted regulations on firearms and ammunition, and three 

have preempted anti-discrimination laws. Id. at 3, 23. 

 
23 Kenneth A. Stahl, Preemption, Federalism, and Local Democracy, 44 FORDHAM URB. L.J. 133, 134 

(2017). For an extensive survey of state preemption of municipal laws, see NAT’L LEAGUE OF CITIES, supra 

note 22; BRIFFAULT ET AL., supra note 4. 

 
24 See Stahl, supra note 23, at 136–43. 

 
25 See Lori Riverstone-Newell, The Rise of State Preemption Laws in Response to Local Policy Innovation, 

47 PUBLIUS: J. FEDERALISM 403, 405 (2017). 

 
26 See, e.g., Jennifer L. Pomeranz & Mark Pertschuk, State Preemption: A Significant and Quiet Threat to 

Public Health in the United States, 107 AM. J. PUB. HEALTH 900 (2017); Jennifer Karas Montez, 

Deregulation, Devolution, and State Preemption Laws’ Impact on U.S. Mortality Trends, 107 AM. J. PUB. 

HEALTH 1749 (2017); Andrew Gillum, Bill Peduto, & Ted Wheeler, Mayors Want to Pass Gun Safety Laws, 

but the NRA and Our State Legislatures Won’t Let Us, USA TODAY (Mar. 23, 2018), 

https://usatoday.com/story/opinion/2018/03/23/mayors-want-gun-control-but-blocked-nra-preemption-laws-

column/450893002/ [https://perma.cc/B372-LYEZ]. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         211 

Of most significance for the purposes of this Article are the three states—North 

Carolina, Tennessee, and Arkansas—that have explicitly preempted municipal legislation 

intended to extend local non-discrimination protections to LGBTQ individuals.27 

Tennessee was the first to do so in 2011.28 Arkansas followed suit in 2015 with the 

Intrastate Commerce Improvement Act (Act 137), which will be discussed in detail in 

Parts II and III. North Carolina is the most recent state to preempt a local NDO, and the 

most high-profile.29 North Carolina’s is also the only law of the three that has been 

repealed, although only in part. 

 

North Carolina’s bill, the Public Facilities Privacy and Security Act, better known as 

HB2, was passed in response to the city of Charlotte’s Ordinance 7056, which prohibited 

discrimination on the basis of sexual orientation and gender identity. HB2 limited “sex” 

to mean the designation of male or female on one’s birth certificate and provided that any 

municipal NDO would be preempted by state regulation.30 

 

The Tennessee bill, called the Equal Access to Intrastate Commerce Act, was a 2011 

law that amended Tennessee’s Human Rights Act to define “sex” as the designation of a 

person as male or female on their birth certificate and prohibited any local government 

from enacting ordinances that create any anti-discrimination practices not covered by 

 
27 See NAT’L LEAGUE OF CITIES, supra note 22, at 10. 

 
28 Equal Access to Intrastate Commerce Act, TENN. CODE ANN. § 7-51-1802 (2017) (passed as House Bill 

600 during the 2011 Session of the 107th General Assembly and effective as codified April 7, 2017). 

 
29 See Paul J. Weber & Will Weissert, ‘Bathroom Bill’ Dies Again in Texas as Session Abruptly Ends, USA 

TODAY (Aug. 16, 2017), https://www.usatoday.com/story/news/politics/2017/08/16/texas-bathroom-

bill/571671001/ [https://perma.cc/W6AL-BYY8]; Amber Phillips, The Tumultuous History of North 

Carolina’s Bathroom Bill, Which Is On Its Way To Repeal, WASH. POST (Mar. 30, 2017), 

https://washingtonpost.com/news/the-fix/wp/2016/12/19/the-tumultuous-recent-history-of-north-carolinas-

bathroom-bill-which-could-be-repealed/?noredirect=on&utm_term=.8e291a062d17/ [https://perma.cc/K6R8-

582B]. 

 
30 Public Facilities Privacy & Security Act, N.C. GEN. STAT. § 143-760 (2016), repealed by 2017 N.C. Sess. 

Laws 1, ch.4. HB2 was later repealed in response to widespread public outrage, but the replacement bill 

retains the preemptive element of HB2 and North Carolina cities remain unable to extend non-discrimination 

protections beyond what is protected at the state level. See This HB2 “Repeal” is Fake, LAMBDA LEGAL 

(Mar. 30, 2017), https://lambdalegal.org/blog/20170330_fake-hb2-repeal [https://perma.cc/89JK-WE3D]. 



 

212 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

state law.31 The law was proposed in response to Nashville’s passage of a city-wide NDO 

that prohibited discrimination on the basis of sexual orientation and gender identity.32 

 

Act 137, Arkansas’ bill, is similar to Tennessee’s. Like the Equal Access to Intrastate 

Commerce Act, Act 137 prohibits cities and counties in Arkansas from adopting 

ordinances or policies that “create[] a protected classification or prohibit[] discrimination 

on a basis not contained in state law.”33 But unlike Tennessee’s bill, Act 137 does not 

offer a definition of “sex.” 

 

It is significant that, though their function was to remove protections for LGBTQ 

individuals, both the Arkansas and Tennessee laws were titled as regulations of intrastate 

commerce. This has allowed both laws to present as facially neutral, and so avoid 

successful federal constitutional challenges34 under the Equal Protection Clause.35 This is 

not to say that there are no constitutional challenges to be mounted, especially for these 

cases which implicate the rights of LGBTQ individuals. As discussed in the next section, 

despite their facial neutrality and purported secular aims, some of these laws are arguably 

impermissibly religiously motivated and so could be found invalid under the 

Establishment Clause of the Constitution. 

 

 

 
31 Equal Access to Intrastate Commerce Act, TENN. CODE ANN. § 7-51-1802 (2017). 

 
32 See Leslie Fenton, The Anti-Gay Tennessee Bill No One’s Talking About, SALON (May 26, 2011), 

https://www.salon.com/2011/05/26/tennessee_antigay_bill_open2011/ [https://perma.cc/CV55-6GPV]. 

 
33 ARK. CODE ANN. §§ 14-1-403 (2015). 

 
34 There are also state constitutional challenges that could potentially be raised against these preemption bills; 

however, they are outside the scope of this Article. For a discussion of state constitutional challenges 

regarding home rule immunity, “generality” requirements, and procedural deficiencies, see BRIFFAULT ET AL., 

supra note 4, at 11–13. 

 
35 U.S. CONST. amend. XIV. The Fourteenth Amendment provides that no state shall “deny to any person 

within its jurisdiction the equal protection of the laws.” Id.; see also Romer v. Evans, 517 U.S. 620 (1996). 

Romer concerned a challenge to a Colorado state referendum which “preclude[d] all legislative, executive, or 

judicial action at any level of state or local government designed to protect the status of persons based on 

their homosexual, lesbian or bisexual orientation, conduct, practices or relationships.” Romer, 517 U.S. at 

620. The law was passed in response to the passage of local ordinances in three Colorado cities that extended 

anti-discrimination protections to individuals on the basis of sexual orientation. Id. The Supreme Court held 

that the state law was unconstitutional under the Equal Protection Clause, on the grounds that the referendum 

“raise[d] the inevitable inference that the disadvantage imposed is born of animosity toward the class of 

persons affected.” Id. at 634. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         213 

B. What Can Cities Do When Faced with Preemption? 

 

In most states, cities are left with few options when it comes to challenging state laws 

that preempt local legislation. The doctrine of preemption refers to the idea that the law 

of a lower authority will be displaced by the law of a higher authority should the two 

come into conflict. Preemption can occur both at the federal36 and state37 level, and it 

generally takes one of two forms: express or implied. Express preemption is created when 

a statute explicitly indicates legislative (state or federal) intent to displace other laws by 

lower authorities.38 Implied preemption exists where the federal or state regulatory 

regime is so comprehensive that it has “fully occupied” the field, such that regulation by 

a lesser authority is intended to be preempted, even if the legislature has not said so 

 
36 Federal preemption is the invalidation of state or municipal law that conflicts with federal law. The source 

of federal preemption is the Supremacy Clause in Article IV of the U.S. Constitution. U.S. CONST. art. IV, cl. 

2. See also Altria Group, Inc. v. Good, 555 U.S. 70, 76 (2008) (“[W]e have long recognized that state laws 

that conflict with federal laws are without effect.”). 

 
37 The sources of intrastate preemption are varied, and there is a lot of diversity in state-local relationships. 

Often, the authority for state preemption takes the form of a grant of home rule authority to cities, either 

through the state constitution or via statute. See, e.g., N.Y. CONST. art. IX (“[E]very local government shall 

have power to adopt and amend local laws not inconsistent with the provisions of this constitution or any 

general law relating to its property, affairs or government.”); ARK. CODE ANN § 14-55-102 (1875) 

(“Municipal corporations shall have power to make and publish bylaws and ordinances, not inconsistent with 

the laws of this state.”). 

 
38 See, e.g., Equal Access to Intrastate Commerce Act, TENN. CODE ANN. §7-51-1802 (2017) (expressly 

displacing a Nashville, Tennessee, ordinance which added sexual orientation and gender identity to the 

classes of persons protected by equal opportunity provisions). The Equal Access to Intrastate Commerce Act 

prohibits municipalities from creating any non-discrimination protections that are not covered by statewide 

anti-discrimination acts. Id. It reads, in part: 

 

No local government shall by ordinance, resolution, or any other means impose on or 

make applicable to any person an anti-discrimination practice, standard, definition, or 

provision that shall deviate from, modify, supplement, add to, change, or vary in any 

manner from . . . the definition of “discriminatory practices” in § 4-21-102. Id. 



 

214 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

explicitly.39 Notably, implied preemption doctrine can displace state or local law even 

absent any actual conflict with federal or state legislation.40 

 

State preemption of municipal law is also impacted by the extent to which the state 

delegates legislative authority to its municipalities. The federal Constitution makes no 

mention of local governments and the Tenth Amendment leaves to the states the authority 

to delegate powers to its cities.41 As a result, there is no real uniformity among states 

regarding state-municipal relations. However, states can generally be split into two 

categories: Dillon’s Rule states and home rule states. Dillon’s Rule42 advances a very 

narrow scope of municipal authority.43 In Dillon’s Rule states, municipalities have only 

powers expressly granted to them by the state, those that are necessarily incidental to 

granted powers, and those that are indispensable to municipal governance.44 Historically, 

 
39 See, for instance, O’Connell v. Stockton, 162 P.3d 583, 587 (Cal. 2007), which states:  

 

When the Legislature has not expressly stated its intent to occupy an area of law, we look 

to whether it has impliedly done so. This occurs in three situations: when (1) the subject 

matter has been so fully and completely covered by general law as to clearly indicate that 

it has become exclusively a matter of state concern; (2) the subject matter has been 

partially covered by general law couched in such terms as to indicate clearly that a 

paramount state concern will not tolerate further or additional local action; or (3) the 

subject matter has been partially covered by general law, and the subject is of such a 

nature that the adverse effect of a local ordinance on the transient citizens of the state 

outweighs the possible benefit to the locality (internal quotations and citations omitted). 

 
40 See Viet D. Dinh, Reassessing the Law of Preemption, 88 GEO. L.J. 2085, 2103 (2000). 

 
41 U.S. CONST. art. X (“The powers not delegated to the United States by the Constitution, nor prohibited by it 

to the States, are reserved to the States respectively, or to the people.”). 

 
42 Dillon’s Rule is named for John F. Dillon, a Justice on the Iowa Supreme Court (1864–69), who expressed 

what would come to be known eponymously as Dillon’s Rule in City of Clinton v. Cedar Rapids & M.R.R. 

Co., 24 Iowa 455, 475 (1868) (“Municipal corporations owe their origin to, and derive their powers and 

rights wholly from, the legislature. It breathes into them the breath of life, without which they cannot exist. 

As it creates, so it may destroy. If it may destroy, it may abridge and control.”). Dillon’s conception was 

endorsed by the Supreme Court in Hunter v. City of Pittsburgh, 207 U.S. 161 (1907), which held that there is 

no right under the federal Constitution to form local governments and that municipalities are merely 

“political subdivisions” of the state. Id. at 178. 

 
43 See Diller, supra note 7, at 1122. 

 
44 See JOHN F. DILLON, TREATISE ON THE LAW OF MUNICIPAL CORPORATIONS § 9(b) (1873). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         215 

Dillon’s Rule was widely accepted, but today only eight states maintain Dillon’s Rule in 

some form.45 

 

Most states moved away from Dillon’s Rule and adopted “home rule” provisions in 

the late nineteenth and early twentieth centuries.46 Home rule allows municipalities to 

exercise any delegable power that the state has not expressly reserved for itself and it can 

be granted via a variety of mechanisms.47 States vary as to the extent of these powers they 

allow their cities to exercise, but almost every state has some version of home rule that is 

operative.48 Forty states provide for home rule either in their state constitutions or 

through statute, though there are differences in its features. For example, some states, like 

Arizona and Illinois, limit home rule authority to cities with certain population sizes.49 

Others, like Delaware, limit home rule based on subject matter—for instance, restricting 

the authority of municipalities to pass laws that create felonies.50 On the whole, no state 

reserves all authority for itself, nor does any state delegate all legislative power to its 

cities, the result being that most states engage some aspects of both Dillon’s and home 

rule.51 

 
45 These states are Alabama, California, Colorado, Illinois, Indiana, Kansas, Louisiana, and Tennessee. In 

almost all these states, however, Dillon’s rule is maintained only for a certain category of local governments, 

such as townships only in Indiana and non-charter cities in California. See AM. CITY CTY. EXCHANGE, 

FEDERALISM, DILLON RULE, AND HOME RULE 5 (2016). 

 
46 See NAT’L LEAGUE OF CITIES, supra note 22, at 5. 

 
47 See Diller, supra note 7, at 1114. Home rule has two general features: initiative and immunity. See 

BRIFFAULT ET AL., supra note 4, at 2. Initiative power, which is widely accepted throughout the United States 

with “at least 40 states delegating some significant, presumptive authority to local governments,” id. at 2, 

“enable[s] local governments to undertake actions over a range of important issues without having to run to 

the state for specific authorization.” RICHARD BRIFFAULT & LAURIE REYNOLDS, CASES AND MATERIALS ON 

STATE AND LOCAL GOVERNMENT LAW 346 (2016). Immunity, by contrast, “protect[s] local government 

decisions concerning local action from displacement by state law.” Id. 

 
48 Forty-four states have some version of home rule that is operative. As with the states that still use Dillon’s 

Rule, many home rule states subject their home rule charters to limitations. In Arizona, for example, home 

rule only applies to cities with populations of at least 35,000; all smaller cities are governed by Dillon’s Rule. 

See AM. CITY CTY. EXCHANGE, supra note 45, at 6. 

 
49 See id. 

 
50 See BRIFFAULT ET AL., supra note 4, at 2. 

 
51 See generally JESSE J. RICHARDSON, JR., MEGHAN ZIMMERMAN GOUGH, & ROBERT PUENTES, BROOKINGS 

INST., IS HOME RULE THE ANSWER?: CLARIFYING THE INFLUENCE OF DILLON’S RULE ON GROWTH 

MANAGEMENT (2003). 



 

216 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

The distinctions between primarily Dillon’s Rule and primarily home rule 

jurisdictions and, more often, distinctions between different home rule systems have 

implications for how empowered cities are to challenge preemption. In states chiefly 

governed by Dillon’s Rule, cities are broadly preempted from legislating in most policy 

areas.52 In pure “legislative” home rule states (the majority rule),53 municipal legislative 

power can be preempted by clear legislative intent.54 For other forms of home rule, 

however, the path to preemption is less clear. 

 

Not all home rule regimes fit the legislative home rule model, and some states have 

home rule regimes that grant immunity to cities against state preemption of certain 

categories of legislation or certain types of municipal action.55 Other systems place 

requirements on preemptory laws themselves, such as the need to have general and 

uniform applicability.56 These provisions create some protections for cities against state 

interference, but they are not universal and most cities remain subject to legislative home 

rule. Further, courts typically default against city power when cities attempt to push the 

boundaries of their authority vis-à-vis states.57 In cases where state and local governments 

conflict, courts often rule in favor of state arguments for consistency, perhaps due to a 

“general judicial uneasiness with creative local action and corresponding preference for 

uniformity.”58 

 

The upshot of preemption regimes for municipalities is that cities have few avenues 

for judicial recourse against state preemptive action, whether or not home rule applies.59 

Local governments do not possess any federalism rights against states under the U.S. 

Constitution and the bar that must be cleared for preemption laws to raise other federal 

 
52 See NAT’L LEAGUE OF CITIES, supra note 22, at 5. 

 
53 See Diller, supra note 7, at 1126. 

 
54 See BRIFFAULT ET AL., supra note 4, at 4. 

 
55 See id. 

 
56 See id. at 5. 

 
57 See Stahl, supra note 23, at 171. 

 
58 David J. Barron, Reclaiming Home Rule, 116 HARV. L. REV. 2257, 2347–48 (2003). 

 
59 See Stahl, supra note 23, at 172. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         217 

constitutional concerns, as will become clear in Sections II and III, is lofty.60 In the 

absence of proving that state laws run afoul of the U.S. Constitution, municipalities are 

left essentially without federal protections against state arrogation of their lawmaking 

authority. These limitations on municipal options mean that federal constitutional claims 

can be a significant part of the arsenal for challenging preemption. One such claim will 

be considered in detail in Parts II and III. 

 

II. The Legal Framework for Federal Constitutional Challenges 

 

 Most of the litigation challenging the laws out of Arkansas, North Carolina, and 

Tennessee have been based on Equal Protection theories. For reasons explained briefly 

below, Equal Protection claims have not had much success in striking down these 

preemptive bills. Especially in the case of Act 137, which on its face makes no mention 

of LGBT populations, an Equal Protection claim is not a probable source of relief. 

However, the particular facts of Act 137’s passage and its legislative history implicate 

religious concerns such that the Establishment Clause could provide grounds for 

invalidating the law where Equal Protection cannot. The likelihood of success depends, in 

part, on which Establishment Clause test the Supreme Court would choose to employ, 

which is discussed in detail below. 

 

A. Equal Protection Challenge 

 

To prevail on an equal protection claim against state preemption laws, cities must 

show that the state law in question either intentionally discriminated against a protected 

 
60 See Hunter v. City of Pittsburgh, 207 U.S. 161, 178–79 (1907). 



 

218 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

class,61 which would trigger the courts to apply heightened scrutiny,62 or that the statute 

was motivated by a bare desire to harm a group and cannot be supported by any rational 

basis.63 Looking at the three state preemption bills that have targeted LGBT non-

discrimination, both North Carolina’s and Tennessee’s expressly limit the definition of 

“sex.” North Carolina limits “sex” to “biological sex”64 and Tennessee limits it to “the 

designation of an individual person as male or female as indicated on the individual’s 

birth certificate.”65 Consequently, it could be argued that these provisions, particularly 

North Carolina’s limitation to “biological sex,” create facial classifications that may 

trigger heightened scrutiny for equal protection purposes.66 

 

Arkansas’ Act 137, in contrast, is facially neutral, and offers no definition of sex or 

any other protected classifications, and so presents a weaker case for an equal protection 

claim.67 The bill provides that municipalities cannot adopt NDOs that create protections 

 
61 Sexual orientation and gender identity are not currently protected classes in federal law. But some lower 

courts have read the protection of sexual orientation and gender identity into protections for gender that exist 

in federal statutes, specifically in the context of Title VII employment claims. See, e.g., Zarda v. Altitude 

Express, Inc., 883 F.3d 100 (2d Cir. 2018) (holding that sexual orientation is a subset of sex discrimination 

for the purposes of Title VII); Equal Emp’t Opportunity Comm’n v. R.G. & G.R. Harris Funeral Homes, Inc., 

884 F.3d 560 (6th Cir. 2018) (finding that the decision to fire an employee who had decided to transition 

from male to female constituted discrimination based on failure to conform to gender norms and was thus a 

form of impermissible sex stereotyping under Title VII); Hively v. Ivy Tech Cmty. Coll. of Ind., 853 F.3d 

339 (7th Cir. 2017) (holding that discrimination based on sexual orientation is a form of sex discrimination 

and is thus prohibited by Title VII of the Civil Rights Act of 1964). The question of whether sex 

discrimination includes discrimination on the basis of sexual orientation and discrimination against 

transgender or gender non-conforming persons has arisen primarily in disputes regarding Titles VII and IX of 

the Civil Rights Act, and until the Supreme Court grants certiorari to resolve the disagreements among the 

lower courts, the law will remain unsettled. 

 
62 To invoke heightened scrutiny, it must be shown that the action in question has a facial classification or is 

motivated by an intent to discriminate; it is not enough to show that a particular group is disparately 

impacted. See Washington v. Davis, 426 U.S. 229, 242 (1976). 

 
63 See Romer v. Evans, 517 U.S. 620, 634–35 (1996). 

 
64 Public Facilities Privacy and Security Act, 2016 N.C. Sess. Laws 3 §§ 1–3, repealed by 2017 N.C. Sess. 

Laws 4 § 1. 

 
65 Equal Access to Interstate Commerce Act, TENN. CODE ANN. § 7-51-1802 (2017). 

 
66 See BRIFFAULT ET AL., supra note 4, at 14. 

 
67 Intrastate Commerce Improvement Act, ARK. CODE ANN. §§ 14-1-402, 14-1-403 (2015). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         219 

not available in state law, including those for sexual orientation and gender identity.68 

Unlike the Tennessee and North Carolina laws, however, it provides no instruction as to 

how sexual orientation and/or gender identity should be determined for purposes of state 

law. The fact that Act 137 does not identify LBGT individuals as targets or explicitly 

reference LGBT individuals in any way presents a barrier to mounting an equal 

protection claim against the bill. 

 

B. Why the Establishment Clause?  

 

Even though it is facially neutral, if Act 137 was motivated by a desire to allow 

business owners to discriminate against LGBT persons on the basis of religious beliefs, it 

could violate the Establishment Clause. The context and circumstances surrounding the 

bill’s passage, especially statements made on the floor of the Arkansas General Assembly 

during debate, suggest that religion played a central role in the decision-making of many 

of the lawmakers in support of the bill.69 Further, only weeks after the passage of Act 

137, Senator Bart Hester and Representative Bob Ballinger—the primary sponsors of Act 

137—also sponsored HB 1228, the Religious Freedom Restoration Act.70 Senator Hester 

has a documented history of being anti-LGBT and has on numerous occasions linked his 

views on homosexuality to his religion.71 In a speech on the steps of the Arkansas state 

capital in 2014, Senator Hester said:  

 

Because God loves each and every one of us equally, he provided us an 

instruction book on how to live our lives . . . and in that instruction book 

for life He says, “because I love you and I want what is best for you I 

think marriage is between one man and one woman.” Friends, that is 

truth.72  

 
68 Id. 

 
69 See infra Part III.A.1. for a more thorough discussion of the statements made during floor debates of Act 

137. 

 
70 See Jessica Glenza, Arkansas Passes Indiana-Style ‘Religious Freedom’ Bill Criticized as Anti-Gay, THE 

GUARDIAN (Mar. 31, 2015), https://www.theguardian.com/us-news/2015/mar/31/arkansas-passes-indiana-

style-religious-freedom-bill [https://perma.cc/WYB2-NNG9]. 

 
71 See Max Brantley, Senate Committee Defeats Ballinger’s Anti-Gay Bill, Also a Sharia Law Attack, ARK. 

TIMES (Feb. 25, 2015), https://www.arktimes.com/ArkansasBlog/archives/2015/02/25/senate-committee-

defeats-ballingers-anti-gay-bill-also-a-shari [https://perma.cc/BZ2Z-CEMS]. 

 
72 ArkansasTeaParty, Rally for True Marriage Senator Bart Hester, YOUTUBE (Nov. 22, 2014), 

https://www.youtube.com/watch?v=PJGh5H3QxlI [https://perma.cc/39MF-A34S]. 



 

220 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

 

Representative Ballinger has a similar history. For example, after Ordinance 119—

Fayetteville’s original bill that extended non-discrimination protection to LGBT 

citizens—was repealed, Ballinger tweeted “Glory to God!”73 On the whole of the facts 

surrounding Act 137, and in the context of a larger national movement of conservative 

and fundamentalist Christian opposition to LGBT rights,74 there is ample reason to 

suggest that the Establishment Clause is an appropriate framework through which to 

challenge Act 137 and legislation like it. 

 

C. Understanding the Supreme Court’s Establishment Clause Precedents 

 

Both the difficulty and the advantage of bringing an Establishment Clause challenge 

is the confusion surrounding the Supreme Court’s decisions on the subject. Both scholars 

and Justices alike have lamented the lack of uniformity and predictability in the Court’s 

Establishment Clause precedents. In a foreword written for a 1990 edition of the Journal 

of Law and Religion, Justice Sandra Day O’Connor discussed the Establishment Clause’s 

“torturous history” in the Court.75 In his dissent in Edwards v. Aguillard, Justice Scalia 

wrote that “[the Court’s] cases interpreting and applying the [Establishment Clause] have 

made such a maze . . . that even the most conscientious governmental officials can only 

guess what motives will be held unconstitutional.”76 The only apparent consensus on the 

Supreme Court’s Establishment Clause jurisprudence is that the Supreme Court has 

reached no consensus on its Establishment Clause jurisprudence. 

 

 

 

 
73 The original tweet has since been deleted. See David Koon, What Will Eureka Do?, ARK. TIMES (Apr. 23, 

2015), https://www.arktimes.com/arkansas/what-will-eureka-do/Content?oid=3823746 

[https://perma.cc/CFF8-5GFK]; Mark Gao, What Everybody Missed During the Fight Over Religious 

Freedom Laws This Year, WASH. POST (Apr. 6, 2015), 

https://www.washingtonpost.com/blogs/govbeat/wp/2015/04/06/what-everybody-missed-during-the-fight-

over-religious-freedom-laws-this-year/?utm_term=.b196d2523c70 [https://perma.cc/2Y69-4TK8]. 

 
74 See, e.g., Amy L. Stone, The Impact of Anti-Gay Politics on the LGBTQ Movement, 10 SOC. COMPASS 459 

(2016); Masha Gessen, How Trump Uses ‘Religious Liberty’ to Attack L.G.B.T. Rights, NEW YORKER (Oct. 

11, 2017), https://www.newyorker.com/news/news-desk/how-trump-uses-religious-liberty-to-attack-lgbt-

rights [https://perma.cc/5P29-4XDX]. 

 
75 Hon. Sandra Day O’Connor, Foreword: The Establishment Clause and Endorsement of Religion, 8 J.L. & 

RELIG. 1, 1 (1990). 

 
76 Edwards v. Aguillard, 107 S. Ct. 2573, 2605 (1987) (Scalia, J., dissenting). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         221 

1. Which Establishment Clause Test Applies?  

 

The Supreme Court has employed at least four different methods of evaluating 

Establishment Clause claims: the Lemon test,77 the endorsement test (which essentially 

functions as a supplement to the Lemon test), the coercion test, and a historically focused 

approach. The Lemon test is three-pronged and inquires as to a law’s purpose, effect, and 

likelihood of creating a government entanglement with religion. Both the Lemon and 

endorsement tests will be discussed in greater detail in the next section. 

 

The coercion test has been proposed by some conservative Justices, such as Scalia 

and Thomas, as a replacement for the Lemon test. It is most clearly articulated in Justice 

Kennedy’s opinion in Lee v. Weisman.78 The Court in Weisman held that the practice of 

public schools inviting clergy members to deliver benedictions violated the Establishment 

Clause because “the school district’s supervision and control of a high school graduation 

ceremony places public pressure, as well as peer pressure, on attending students to stand 

as a group . . . This pressure, though subtle and indirect, can be as real as any other 

compulsion.”79 While Justices have disagreed about the scope of the coercion test—

Justice Scalia, for instance, favored a far narrower definition of coercion than was relied 

upon by Justice Kennedy in Weisman—the crux of the coercion standard is that the 

government does not violate the Establishment Clause unless it provides direct aid to 

religion in such a way as would establish a state church or coerces individuals to support 

or participate in a religion against their will. The coercion test has been utilized by the 

Court’s majority in a few cases but has not been widely adopted.80 

 

On a few occasions, the Court has eschewed formal “tests” altogether and looked 

instead to historical practice. Most notably in Marsh v. Chambers, the Court upheld a 

practice of the Nebraska Legislature to open legislative sessions in prayer led by a state-

paid chaplain.81 Relying on the fact that legislative prayer has been in practice in the 

 
77 Lemon v. Kurtzman, 403 U.S. 602 (1971). 

 
78 Lee v. Weisman, 505 U.S. 577 (1992). 

 
79 Id. at 593. 

 
80 See Elk Grove Unified Sch. Dist. v. Newdow, 542 U.S. 1 (2004); Santa Fe Indep. Sch. Dist. v. Doe, 530 

U.S. 290 (2000). For a more thorough consideration of the Supreme Court’s use of the coercion test, see 

Cynthia V. Ward, Coercion and Choice Under the Establishment Clause, 39 U.C. DAVIS L. REV. 1621 

(2006). 

 
81 Marsh v. Chambers, 463 U.S. 783 (1983). 



 

222 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

United States for almost 200 years and that members of the first Congress engaged in the 

practice despite many of them having just written the Establishment Clause, the Court 

held that the principle of non-establishment of religion and legislative prayer could co-

exist.82 Since Marsh, the Court has applied this historical inquiry analysis almost 

exclusively in cases concerning legislative prayer.83 The Court would almost certainly not 

rely on a historical practice standard in evaluating religiously motivated lawmaking of the 

kind considered in this Article. 

 

Because most Establishment Clause cases that have gone to the Supreme Court have 

concerned the relationship between religion and public education84 or public displays of 

religious symbols,85 it is not entirely clear what test the Supreme Court would employ in 

evaluating the validity of a religiously motivated law, but the majority of the Court’s 

decisions regarding the establishment of religion have prioritized the effects rather than 

their motivations of challenged laws in determining constitutionality. 

 

It is likely, however, that the Supreme Court would apply the Lemon test when 

evaluating Act 137 for two reasons: First, Act 137 does not involve a longstanding 

practice and there is no state coercion. Second, Act 137 involves an issue of discord 

between stated and actual legislative purpose, and when the Court has considered issues 

of that nature in the past, it has used the Lemon test. For example, in McGowan v. 

Maryland, the Court acknowledged that a Maryland law requiring most businesses to 

remain closed on Sundays was “undeniably religious in origin.”86 Nonetheless, it 

concluded that the law did not violate either the Free Exercise or Establishment Clauses 

of the First Amendment because there is secular value in having a uniform day of rest and 

 
82 Id. at 790. 

 
83 See, e.g., Town of Greece v. Galloway, 572 U.S. 565, 591 (2014) (holding that a town’s practice of 

opening board meetings in prayer did not violate the Establishment Clause). 

 
84 See, e.g., Elk Grove Unified Sch. Dist., 542 U.S. at 1 (considering whether the words “under God” in the 

Pledge of Allegiance invalidated a policy of reciting the pledge daily); Wallace v. Jaffree, 472 U.S. 38 (1985) 

(concerning a daily moment of silence for meditation and voluntary prayer in a public school); Engel v. 

Vitale, 370 U.S. 421 (1962) (regarding the recitation of a daily morning prayer in a public school); Everson v. 

Bd. of Ed. of Ewing Twp., 330 U.S. 1 (1947) (concerning the provision of reimbursements to parents who 

used publicly operated buses to send children to parochial schools). 

 
85 See, e.g., Van Orden v. Perry, 545 U.S. 677 (2005) (concerning a monument to the Ten Commandments 

displayed in the Texas state capitol); Cty. of Allegheny v. ACLU Greater Pittsburgh Chapter, 492 U.S. 573 

(1989) (considering the display of a nativity scene in a county’s public holiday display); Lynch v. Donnelly, 

465 U.S. 668 (1984) (considering the display of a nativity scene in a Christmas display in a public park). 

 
86 McGowan v. Maryland, 366 U.S. 420, 446 (1961). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         223 

most citizens, given the opportunity to do so, would choose Sunday as that day of their 

own accord.87 Though decided before Lemon, the Court’s analysis in McGowan 

demonstrates the origins of what would become the Lemon test. 

 

The Court has, in some cases, struck down religious laws on purpose grounds. In 

Stone v. Graham, the Court considered a Kentucky statute that required the posting of the 

Ten Commandments on the walls of public school classrooms.88 Applying the Lemon 

test, the Court found that the statute was invalid for failing to demonstrate a valid secular 

purpose.89 Both McGowan and Stone considered issues similar to those created by the 

preemptive bills out of Arkansas, Tennessee, and North Carolina. Both cases presented 

questions of the validity of statutes that were obviously influenced by religion. In both 

cases, the Court applied the Lemon test (or a comparable inquiry) to determine 

constitutionality. Based on these cases, it is more likely than not that the Court would 

choose to apply the Lemon test if presented with a case like Act 137 in Arkansas. 

 

2. Interpreting the Lemon Test 

 

The Lemon test remains the primary vehicle through which the Court evaluates 

Establishment Clause claims despite its utilization of other tests at times. Established in 

Lemon v. Kurtzman, the test establishes three requirements for the government’s 

engagement with religion: (1) government actions must have a secular legislative 

purpose, (2) the government action’s principal or primary effect must neither inhibit nor 

advance religion, and (3) the government action must not foster an “excessive 

government entanglement with religion.”90 If a government action fails any one of these 

prongs, it is unconstitutional under the Establishment Clause.91 

 

a. Secular Purpose 

 

The secular purpose prong of the Lemon test seeks to ensure that legislatures do not 

pass overtly religious legislation. The difficulty is that overtly religious legislation is 

extremely rare, so the question becomes how to distinguish legislation that seeks 

 
87 Id. at 452. 

 
88 Stone v. Graham, 449 U.S. 39 (1980). 

 
89 Id. at 42–43. 

 
90 Lemon v. Kurtzman, 403 U.S. 602, 612–13 (1971). 

 
91 Id. 



 

224 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

impermissible religious ends from legislation that merely “happens to coincide or 

harmonize with the tenets of some or all religions.”92 The secular purpose prong of the 

Lemon test may provide a means of invalidating statutes based on religious motivation, 

but the Court has been somewhat inconsistent in the role that legislative motivation plays 

in its secular purpose analysis. 

 

Generally speaking, the Supreme Court has shown a great deal of deference toward a 

state’s proffered statutory purposes. In Lynch v. Donnelly, the majority noted that “the 

Court has invalidated legislation or governmental action on the ground that a secular 

purpose was lacking, but only when it has concluded there was no question that the 

statute or activity was motivated wholly by religious considerations.”93 The majority in 

Santa Fe Independent School District v. Doe wrote that “[w]hen a governmental entity 

professes a secular purpose for an arguably religious policy, the government’s 

characterization is, of course, entitled to some deference. But it is nonetheless the duty of 

the courts to distinguish a sham secular purpose from a sincere one.”94 Both of these 

cases indicate that the Court often gives deference to stated legislative purpose, but the 

strength of deference has varied. 

 

Even in cases where the benefits to a particular religion were significant, the Court 

has accepted legislatures’ statements of secular purpose and not questioned them 

further.95 For example, in Everson v. Board of Education, the Court upheld a New Jersey 

statute providing reimbursements funded by taxes for the bus fares of children who 

attended both public and parochial schools on the grounds that the legislature’s stated 

purpose of enabling all children to receive a secular education was valid.96 However, this 

deference is not absolute. Over time, the Court has become more receptive to the idea 

that governmental motives can be appropriately investigated and that, in certain 

circumstances, such investigation is necessary to evaluate constitutionality, though the 

 
92 McGowan, 366 U.S. at 442. 

 
93 465 U.S. 668, 680 (1984). 

 
94 530 U.S. 290, 308 (2000). 

 
95 See, e.g., Bd. of Ed. of Cent. Sch. Dist. No. 1 v. Allen, 392 U.S. 236, 248 (1968) (upholding the expressed 

secular purpose of furthering educational purposes to young people for a New York statute requiring school 

districts to provide textbooks to students in both public and religious schools); Walz v. Tax Comm’n of City 

of New York, 397 U.S. 664, 672 (1970) (“The legislative purpose of a property tax exemption is neither the 

advancement nor the inhibition of religion; it is neither sponsorship nor hostility.”). 

 
96 Everson v. Bd. of Ed. of Ewing Twp., 330 U.S. 1 (1947). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         225 

practice has been employed inconsistently.97 In 2005 in McCreary County v. ACLU, for 

example, the Court considered a Kentucky statute requiring that the Ten Commandments 

be posted in courthouses.98 In holding that the statute violated the Establishment Clause, 

the Court relied heavily on an analysis of the legislature’s motivations in enacting the 

law.99 Writing for the majority, Justice Souter emphasized that inquiry into legislative 

purpose is both “a staple of statutory interpretation” and “a key element of constitutional 

doctrine.”100 Furthermore, on the relationship between legislative purpose and the 

Establishment Clause, Justice Souter wrote that “scrutinizing purpose does make 

practical sense, as in Establishment Clause analysis, where an understanding of official 

objective emerges from readily discoverable fact, without any judicial psychoanalysis of 

a drafter’s heart of hearts.”101  

 

One difficulty that the Court’s reasoning in McCreary presents is whether there is, in 

theory or in fact, a distinction between legislative purpose and legislative motivation. 

Despite appeals from scholars to maintain a distinction between the two categories, 

“courts [in practice] often use both terms interchangeably.”102 McCreary is an example of 

this. In one section of the majority opinion, Justice Souter writes that “[t]he eyes that look 

to purpose belong to an ‘objective observer,’ one who takes account of the traditional 

external signs that show up in the ‘text, legislative history, and implementation of the 

statute,’ or comparable official act.”103 Only two paragraphs later, and within the same 

section, he posits that  

 

If someone in the government hides religious motive so well that the 

“objective observer, acquainted with the text, legislative history, and 

implementation of the statute” . . . cannot see it, then without something 

 
97 See Frederick Mark Gedicks, Motivation, Rationality, and Secular Purpose in Establishment Clause 

Review, 1985 ARIZ. ST. L.J. 677, 687 (1985). 

 
98 McCreary Cty. v. ACLU of Ky., 545 U.S. 844 (2005). 

 
99 Id. at 861. 

 
100 Id. 

 
101 Id. at 862. 

 
102 Robert C. Farrell, Legislative Purpose and Equal Protection’s Rationality Review, 37 VILL. L. REV. 2, 5 

(1992). 

 
103 McCreary, 545 U.S. at 862. 



 

226 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

more the government does not make a decisive announcement that in 

itself amounts to taking religious sides.104 

 

Essentially, Justice Souter acknowledges that the same evidentiary sources—text, 

legislative history, and implementation—serve inquiries of both purpose and motivation, 

suggesting that the two categories are difficult to sever. 

 

In another Establishment Clause case decided twenty years before McCreary, 

Wallace v. Jaffree, Justice Stevens wrote on behalf of the majority that government 

action is invalid under the First Amendment “if it is entirely motivated by a purpose to 

advance religion.”105 Wallace concerned a challenge to an Alabama statute that 

authorized a moment of silence in public schools.106 In invalidating the statute, the Court 

found that “the record not only provides us with an unambiguous affirmative answer, but 

it also reveals that the passage of [the statute] was not motivated by any clearly secular 

purpose—indeed, the statute had no secular purpose.”107 Here, again, the functional 

distinction between motivation and purpose is blurred. It appears that, especially in cases 

where inquiring as to religious motive does not, to parrot Judge Souter’s language, 

require delving into a legislator’s “heart of hearts,” legislative motivation and legislative 

purpose are used by the Court to mean roughly the same thing. This seems to be 

particularly true in cases such as McCreary, where the facts establish that legislators wore 

their impermissible motivations on their sleeves, so to speak. 

 

Of course, the inquiry into legislative purpose and motivation for Establishment 

Clause cases depends on the Court’s decision to utilize the Lemon test. Compare 

McCreary to Van Orden v. Perry, a case with almost identical facts to McCreary and 

decided by the Court on the same day but with the opposite outcome.108 In McCreary, the 

Court held that the Kentucky statute requiring the Ten Commandments be posted in 

courthouses was a violation of the Establishment Clause.109 In Van Orden, the same 

 
104 Id. at 863. 

 
105 Jaffree, 472 U.S. at 38. 

 
106 Id. 

 
107 Id. at 56 (emphasis in original). 

 
108 Van Orden v. Perry, 545 U.S. 677 (2005). 

 
109 McCreary, 545 U.S. at 861. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         227 

Court found that a monument to the Ten Commandments displayed on the grounds of the 

Texas State Capitol did not violate the Establishment Clause.110 

 

Two things might explain this somewhat baffling outcome. First is the Court’s own 

explanation for the different holdings, which identifies context as the distinguishing 

factor. In the plurality opinion for Van Orden, Chief Justice Rehnquist characterized the 

location of the Ten Commandments monument on the State Capitol grounds as a “passive 

use of those texts” and noted that this setting “suggests little or nothing of the sacred.”111 

He wrote: 

 

In certain contexts, a display of the tablets of the Ten Commandments 

can convey not simply a religious message but also a secular moral 

message . . . [a]nd . . . in certain contexts, a display of the tablets can also 

convey a historical message (about a historic relation between those 

standards and the law).112 

 

By contrast, the display of the Ten Commandments in courthouses in McCreary lacked 

the historical and secular function of the display in Van Orden. A second, related 

explanation for the different outcomes is the application of the Lemon test in McCreary 

versus the historical inquiry test in Van Orden. The Van Orden majority eschewed the 

Lemon test, turning instead to the historical practice inquiry from Marsh v. Chambers.113 

The McCreary majority, on the other hand, used Lemon’s purpose prong to invalidate the 

display.114  

 

The space created in the Lemon test for the consideration of legislative motives could 

bode well for challenges to religiously motivated preemption laws. On all five occasions 

on which the Supreme Court struck down laws under the secular purpose prong of the 

Lemon test,115 it emphasized context and circumstances. In Edwards v. Aguillard, the 

 
110 Van Orden, 545 U.S. at 681. 

 
111 Id. at 691. 

 
112 Id. at 701. 

 
113 Van Orden, 545 U.S. at 677–78. 

 
114 McCreary, 545 U.S. at 864–65. 

 
115 McCreary Cty. v. ACLU of Ky., 545 U.S. 844 (2005); Santa Fe Indep. Sch. Dist. v. Doe, 530 U.S. 290 

(2000); Edwards v. Aguillard, 107 S. Ct. 2573 (1987); Wallace v. Jaffree, 472 U.S. 38 (1985); Stone v. 

Graham, 449 U.S. 39 (1980) (per curiam). 



 

228 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

Court held unconstitutional a Louisiana law requiring that where evolution was taught in 

public schools, creationism also had to be taught.116 The Court found that the statute was 

primarily intended to advance Christianity, not to “protect academic freedom,” as its 

preamble alleged.117 In reaching this conclusion, the Justices relied on the bill’s 

legislative history, the statements of its sponsors, and the context in which the law was 

passed: the connection between Christianity and anti-evolutionary teaching at large.118 

Justice Brennan, writing for the majority, explained that the Court need not always take at 

face value the proffered purpose of a law, especially when religion is involved: 

 

As in Stone and Abington, we need not be blind in this case to the 

legislature’s preeminent religious purpose in enacting this statute. There 

is a historic and contemporaneous link between the teachings of certain 

religious denominations and the teaching of evolution. It was this link 

that concerned the Court in Epperson v. Arkansas, which also involved a 

facial challenge to a statute regulating the teaching of evolution. In that 

case, the Court reviewed an Arkansas statute that made it unlawful for an 

instructor to teach evolution or to use a textbook that referred to this 

scientific theory. Although the Arkansas antievolution law did not 

explicitly state its predominant religious purpose, the Court could not 

ignore that “[t]he statute was a product of the upsurge of 

‘fundamentalist’ religious fervor” that has long viewed this particular 

scientific theory as contradicting the literal interpretation of the Bible.119 

 

Justice Brennan acknowledged that circumstances beyond the particular facts of a 

case at hand, such as an established link between certain religions and teachings, 

can bear on the Court’s assessment of the motivations of a particular legislature. 

 

Ultimately, the first prong of the Lemon test is the most likely ground on 

which an Establishment Clause challenge to religiously motivated preemption of 

municipal non-discrimination law would succeed. This would depend, however, 

on two things: (1) the Court’s decision to apply the Lemon test as opposed to one 

of its other tests, which are explained below, and (2) a more than cursory inquiry 

into legislative motivation under the Lemon test’s first prong. The likelihood of 

 
116 Aguillard, 107 S. Ct. at 2584. 

 
117 Id. at 2580. 

 
118 Id. at 2577–84. 

 
119 Id. at 2580–81 (internal citations omitted). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         229 

the Court deciding to inquire beyond Act 137’s stated secular purpose is low, but 

certainly possible, and will be discussed in greater detail in Section III. 

 

b. Principal or Primary Effect  

 

The second prong of the Lemon test is the requirement that the principal or primary 

effect of a legislative action must neither inhibit nor advance religion.120 Under this 

prong, a law can be invalidated irrespective of legislative intent if the impact of the action 

violates the principle of government neutrality with respect to religion.121 Traditionally, 

the Court has upheld laws having only a trivial or incidental effect of advancing 

religion.122 For example, in Widmar v. Vincent the Court held that the University of 

Missouri’s refusal to allow a religious student organization to have worship services in its 

student center was a violation of the students’ First Amendment rights, and any benefit 

the group would derive from the University’s accommodation would not be sufficient to 

violate the Establishment Clause.123 The Court reiterated this idea in Mueller v. Allen, 

stating that “any program which in some manner aids an institution with a religious 

affiliation” does not per se violate the Establishment Clause.124 

 

To fail the principal or primary purpose prong, a law must go beyond merely 

conferring a benefit on a particular religion. To guide this inquiry, courts often look to 

whether a law violates one of the “three main evils against which the Establishment 

Clause was intended to afford protection: sponsorship, financial support, and active 

involvement of the sovereign in religious activity.”125 It is not sufficient to determine 

unconstitutionality to demonstrate that a law provides advantages to religious groups or 

 
120 Lemon, 403 U.S. at 643. 

 
121 See Everson, 330 U.S. at 18 (“[The First Amendment] requires the state to be a neutral in its relations with 

groups of religious believers and non-believers.”). 

 
122 See Carl H. Esbeck, The Lemon Test: Should it be Retained, Reformulated, or Rejected?, 4 NOTRE DAME 

J.L. ETHICS & PUB. POL’Y 513, 521 (1990). 

 
123 454 U.S. 263, 273 (1981) (“It is possible—perhaps even foreseeable—that religious groups will benefit 

from access to University facilities . . . [but] a religious organization’s enjoyment of merely ‘incidental’ 

benefits does not violate the prohibition against the ‘primary advancement’ of religion.”). 

 
124 463 U.S. 388, 393 (1983) (quoting Hunt v. McNair, 413 U.S. 734, 742 (1973)). 

 
125 Lemon, 403 U.S. at 612. 



 

230 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

persons.126 But the Court has, on occasion, struck down laws for failing this prong. In 

Committee for Public Education v. Nyquist, a New York statute that provided grants to 

parochial schools for maintenance and school repairs was held invalid under the 

Establishment Clause.127 Applying the Lemon test, Justice Powell found that the laws had 

the principle effect of promoting religion because “their effect, inevitably, [was] to 

subsidize and advance the religious mission of sectarian schools.”128 Public education 

reimbursement schemes, such as the one in Nyquist, are a common type of law that the 

Court has repeatedly struck down under this prong.129 

 

c. Excessive Entanglement  

 

The third and final prong of the Lemon test asks if the law fosters “an excessive 

government entanglement with religion.”130 According to Chief Justice Burger’s opinion 

in Lemon, determining whether there is an excessive entanglement between government 

and religion requires an examination of “the character and purpose of the institutions that 

are benefitted, the nature of the aid that the State provides, and the resulting relationship 

between the government and the religious authority.”131 The excessive entanglement 

prong is directed at two separate kinds of government involvement with religion: 

administrative and political.132 

 

 
126 See Allen, 392 U.S. at 248 (holding that a New York state law which required public schools to lend 

textbooks to private, including religious, schools did not violate the Establishment Clause). The majority 

wrote: “We cannot agree with appellants either that all teaching in a sectarian school is religious or that the 

processes of secular and religious training are so intertwined that secular textbooks furnished to students by 

the public are in fact instrumental in the teaching of religion.” Id. 

 
127 Comm. for Pub. Ed. & Religious Liberty v. Nyquist, 413 U.S. 756, 797–98 (1973). 

 
128 Id. at 779–80. 

 
129 See, e.g., Sloan v. Lemon, 413 U.S. 825, 835 (1973) (invalidating a law in Pennsylvania that provided 

reimbursement to parents who paid tuition to send their children to nonpublic, usually parochial, schools); 

Aguilar v. Felton, 473 U.S. 402, 414 (1985) (holding that New York City’s use of federal funds to send 

teachers into private religious schools to give remedial instruction violated the Establishment Clause). 

 
130 Lemon, 403 U.S. at 613. 

 
131 Id. at 615. 

 
132 See James A. Serritella, Tangling with Entanglement: Toward a Constitutional Evaluation of Church-

State Contracts, 44 L. & CONTEMP. PROBS. 145 (1981). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         231 

In some cases, the Court has found that a legislative scheme necessitates the creation 

of an ongoing relationship between the government and a religious body or program. In 

Lemon, which concerned statutes in Rhode Island and Pennsylvania that required the 

state to fund some aspects of private parochial schools, the Court held that the ongoing 

supervision and monitoring of these schools by the state to ensure that the conditions of 

the grants were being met would create an excessive entanglement between the 

government and the religious schools that received aid.133 This type of association 

violates the Establishment Clause because it “creates an intimate and continuing 

relationship between church and state.”134 The Court has on a number of occasions 

expressed discomfort with this kind of administrative entanglement.135 

 

Political divisiveness is another standard by which the Court could invalidate a 

statute under the excessive entanglement prong, but it has rarely chosen to do so.136 

Under this doctrine, a statute is invalid if it has the potential to create political division 

along religious lines.137 For example, in Larkin v. Grendel’s Den, the Court held that a 

law which allowed churches to exercise veto power over the grant of liquor licenses to 

establishments within a certain distance of church property violated the Establishment 

Clause.138 Chief Justice Burger wrote for the majority that “[t]he challenged statute thus 

enmeshes churches in the processes of government and creates the danger of ‘[political] 

fragmentation and divisiveness along religious lines.’”139 Some scholars have suggested 

that the political divisiveness test be elevated to the level of other independent 

 
133 Lemon v. Kurtzman, 403 U.S. 602, 624–25 (1971). 

 
134 Id. at 622. 

 
135 See, e.g., Meek v. Pittenger, 421 U.S. 349, 372–73 (1975) (invalidating a textbook loan provision in a 

Pennsylvania statute on the grounds that it would require monitoring of parochial school teachers to ensure 

textbooks were being used for non-religious purposes, and this would create an excessive entanglement of 

government and religion); Levitt v. Comm. for Pub. Ed. & Religious Liberty, 413 U.S. 472, 481–82 (1973) 

(striking down a New York statute providing aid from public funds to nonpublic schools because the 

relationship between state aid and a religious institution could constitute an excessive entanglement). 

 
136 See Lynch v. Donnelly, 465 U.S. 668, 689 (1984) (O’Connor, J., concurring) (“Although several of our 

cases have discussed political divisiveness under the entanglement prong of Lemon . . . we have never relied 

on divisiveness as an independent ground for holding a government practice unconstitutional.”). 

 
137 Lemon, 403 U.S. at 622. 

 
138 459 U.S. 116 (1982). 

 
139 Id. at 127 (quoting Lemon, 403 U.S. at 623). 



 

232 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

Establishment Clause tests,140 but others believe that the doctrine should be abandoned 

entirely.141 Like most of the Court’s Establishment Clause jurisprudence, the status of the 

political divisiveness doctrine is unclear, but it is unlikely that it would be invoked in a 

case involving the standardization of non-discrimination law, such as Act 137. 

 

d. Justice O’Connor’s Endorsement Test 

 

Since Lemon v. Kurtzman was decided, many Justices have expressed desires to 

overturn or significantly reform the Lemon test. Justice Scalia once described the test as 

being “like some ghoul in a late-night horror movie that repeatedly sits up in its grave and 

shuffles abroad, after being repeatedly killed and buried.”142 Justices Thomas and 

Kennedy, too, have expressed desires to replace the Lemon test with another standard of 

evaluating the establishment of religion. One such suggestion for reforming the Lemon 

test that has gained some traction is the so-called “endorsement test” proposed by Justice 

O’Connor in her concurrence in Lynch v. Donnelly.143 

 

Proposed explicitly as an attempt to clarify the Court’s Establishment Clause 

doctrine, Justice O’Connor’s endorsement test suggests that “the proper inquiry under the 

purpose prong of Lemon . . . is whether the government intends to convey a message of 

endorsement or disapproval of religion.”144 For Justice O’Connor, the primary concern of 

an Establishment Clause inquiry should be whether a government action functionally 

creates a religious endorsement which “sends a message to nonadherents that they are 

outsiders, not full members of the political community, and an accompanying message to 

adherents that they are insiders, favored members of the political community.”145 The 

standard for determining if the government has engaged in an endorsement of religion is 

 
140 See, e.g., David R. Schiedemantle, Political Entanglement as an Independent Test of Constitutionality 

Under the Establishment Clause, 52 FORDHAM L. REV. 1209, 1211 (1984). 

 
141 See, e.g., Edward M. Gaffney, Political Divisiveness Along Religious Lines: The Entanglement of the 

Court in Sloppy History and Bad Public Policy, 24 ST. LOUIS U.L.J. 205 (1980). 

 
142 Lamb’s Chapel v. Ctr. Moriches Union Free Sch. Dist., 508 U.S. 384, 398 (1993) (Scalia, J., concurring). 

 
143 465 U.S. 668, 691 (1984) (O’Connor, J., concurring). 

 
144 Id. 

 
145 Id. at 688. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         233 

whether a reasonable observer would think that the activity in question was a 

governmental endorsement of religion.146 

 

In Justice O’Connor’s original conception the endorsement test functioned as a gloss 

on Lemon, refining the first two prongs of the Lemon test to be evaluated through the lens 

of government endorsement.147 According to Justice O’Connor, the utility of the test was 

that it gave content to the otherwise vague and highly subjective inquiries of the purpose 

and effect prongs of the Lemon test.148 Like the Lemon test, the endorsement test has been 

criticized by other members of the Court. But since Justice O’Connor conceived it, it has 

been used in Establishment Clause cases, sometimes as a substitute for Lemon and 

sometimes as a supplement to it.  

 

In 2000, the Court applied the endorsement test (in conjunction with the Lemon test) 

in Santa Fe Independent School District v. Doe, which involved a challenge to Santa Fe 

High School’s practice of beginning football games with a student-led prayer.149 Justice 

Stevens, on behalf of the majority, wrote:  

 

In this context the members of the listening audience must perceive the 

pregame message as a public expression of the views of the majority of 

the student body delivered with the approval of the school 

administration. In cases involving state participation in a religious 

activity, one of the relevant questions is “whether an objective observer, 

acquainted with the text, legislative history, and implementation of the 

statute, would perceive it as a state endorsement of prayer in public 

schools.”150 

 

 
146 See Capitol Square Review & Advisory Bd. v. Pinette, 515 U.S. 753, 777 (1995) (O’Connor, J., 

concurring). 

 
147 See Lynch v. Donnelly, 465 U.S. 668, 687–94 (1984) (O’Connor, J., concurring); Wallace v. Jaffree, 472 

U.S. 38, 67–84 (1985) (O’Connor, J., concurring). 

 
148 See Jaffree at 69–70. 

 
149 530 U.S. 290 (2000). 

 
150 Id. at 308. 



 

234 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

Justice Stevens went on to conclude that the practice was impermissible because it 

communicated to spectators who were non-Christians that they were outsiders, and thus 

violated the Establishment Clause.151  

 

However, more recent cases suggest that the endorsement test is falling out of favor 

with the Court. Elk Grove Unified School District v. Newdow, a 2004 case concerning the 

words “under God” in the Pledge of Allegiance, presented an ideal opportunity to apply 

the endorsement test, but the Court declined to do so, deciding the case instead on a 

standing issue.152 Nor did the court apply the endorsement test in either of the Ten 

Commandments cases of 2005.153 Most recently, in 2014, the Court declined to apply the 

endorsement test in Town of Greece v. Galloway, despite the fact that the Second Circuit 

had expressly applied it in reaching its decision.154 Reversing the Second Circuit, the 

Supreme Court relied on its historical practice standard rather than engage in an 

endorsement analysis. 

 

This shift away from an endorsement standard makes sense given that the test’s 

strongest proponents—Justices Souter, Stevens, and O’Connor—have all retired. If the 

Court is purposefully moving away from the endorsement test, however, it is unclear 

what will replace it. If Justices Kavanaugh and Gorsuch, the most recent conservative 

appointees to the Court, sympathize with their conservative predecessor’s disdain for the 

Lemon test and support a coercive standard replacement, that could be where the Court’s 

Establishment Clause jurisprudence is heading. Should coercion become the new test for 

constitutionality under the Establishment Clause, challenges to laws such as the one 

considered in the following section will become essentially impossible to mount. 

 

 

 

 

 

 
151 Id. at 309. 

 
152 542 U.S. 1 (2004). 

 
153 See McCreary Cty. v. ACLU Union of Ky., 545 U.S. 844 (2005); Van Orden v. Perry, 545 U.S. 677 

(2005). 

 
154 Galloway v. Town of Greece, 681 F.3d 20, 30 (2d Cir. 2012) (“We conclude, on the record before us, that 

the town’s prayer practice must be viewed as an endorsement of a particular religious viewpoint.”), rev’d 

Town of Greece, N.Y. v. Galloway, 572 U.S. 565 (2014). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         235 

III. Act 137 and the Way Forward for the Establishment Clause and Legislative 

Motive  

 

In August of 2014, the City Council of Fayetteville, Arkansas, passed Ordinance 

5703.155 The bill expanded Fayetteville’s nondiscrimination protections to prohibit 

discrimination on the basis of sexual orientation and/or gender identity, specifically in the 

areas of housing, employment, and public accommodation.156 The ordinance’s Purpose 

section read, in relevant part:  

 

The purpose of this chapter is to protect and safeguard the right and 

opportunity of all persons to be free from discrimination based on real or 

perceived race, ethnicity, national origin, age, gender, gender identity, 

gender expression, familial status, marital status, socioeconomic 

background, religion, sexual orientation, disability and veteran status.157 

 

The Ordinance was later repealed by vote and replaced by Ordinance 5781, which was 

substantively similar to Ordinance 5703, but featured a few changes concerning religious 

exemptions and penalties for violation.158 By February 2, 2015, the Intrastate Commerce 

 
155 FAYETTEVILLE, ARK., ORDINANCE 5703, ch. 119 (2014) (repealed Dec. 9, 2014), 

https://www.fayettevilleflyer.com/wp-content/uploads/2014/11/chapter-119-civil-rights-ordinance.pdf 

[https://perma.cc/h9xs-l4lw]. 

 
156 Id. 

 
157 Id. at § 119.01. 

 
158 The ordinance was repealed by voters in December 2014. Many citizens took issue with the wording of 

the ordinance, arguing that it was too vague and non-specific and would open the city up to litigation. See 

City Wire Staff, Fayetteville Voters Approve Repeal of Sexual Orientation Ordinance, TALK BUS. & POL. 

(Dec. 9, 2014), https://talkbusiness.net/2014/12/fayetteville-voters-approve-repeal-of-sexual-orientation-

ordinance/ [https://perma.cc/M4ZZ-A65U]. Others opposed the bill on religious grounds, claiming that it 

threatened the liberty interests of religious citizens and businesses. See Cavan Sieczkowski, Michelle Duggar 

Records Transphobic Robocall Opposing Arkansas Anti-Discrimination Bill, HUFFINGTON POST (Aug. 19, 

2014), https://www.huffingtonpost.com/2014/08/18/michelle-duggar-anti-discrimination-

proposal_n_5689840.html [https://perma.cc/E7A3-C4QF]. The vote to repeal was close, with 52% of 

Fayetteville’s voters in favor. See Todd Gill, Voters Repeal Civil Rights Ordinance in Fayetteville, 

FAYETTEVILLE FLYER (Dec. 9, 2014), https://www.fayettevilleflyer.com/2014/12/09/voters-repeal-civil-

rights-ordinance-in-fayetteville/ [https://perma.cc/KV9X-KVEC]. Despite this repeal, the City Council 

proposed another LGBT non-discrimination ordinance in response to the state legislature’s passage of Act 

137 that offered the same anti-discrimination protections to LGBT individuals but included revised language 

concerning religious exemptions and penalties for violating the ordinance; voters approved the new ordinance 

on September 8, 2015. See CITY OF FAYETTEVILLE LGBT “UNIFORM CIVIL RIGHTS PROTECTION ORDINANCE,” 

ORDINANCE 5781 (SEPTEMBER 2015), https://ballotpedia.org/City_of_Fayetteville_LGBT_%22Uniform_ 



 

236 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

Improvement Act (Act 137) was on the floor of the Arkansas Senate.159 On February 17, 

it became state law.  

 

Act 137 prohibits cities and counties in Arkansas from adopting ordinances or 

policies that “create[] a protected classification or prohibit[] discrimination on a basis not 

contained in state law.”160 The bill’s stated purpose is “to improve intrastate commerce by 

ensuring that businesses, organizations, and employers doing business in the state are 

subject to uniform nondiscrimination laws and obligations.”161 Opponents of the bill 

assert that, regardless of the stated purpose, the Act’s actual purpose is to protect 

discrimination against LGBT people. By the time the Act went into effect on July 22, 

2015, seven Arkansas cities and counties, in addition to Fayetteville, had passed LGBT-

inclusive NDOs in response to the state law.162 

 

Fayetteville attacked the constitutionality of Act 137 in Protect Fayetteville v. City of 

Fayetteville.163 A conservative advocacy group, Protect Fayetteville, brought suit against 

Ordinance 5781 claiming that the law violated Act 137; Fayetteville, in defense of the 

law, argued in part that Act 137 was invalid because it violated the Equal Protection 

 
Civil_Rights_Protection_Ordinance,%22_Ordinance_5781_(September_2015) [https://perma.cc/PLR5-

5ZW8]. The new ordinance was not repealed and is currently being litigated. The Arkansas Supreme Court 

already rejected one argument made by the ordinance’s supporters and remanded the case for consideration 

of the Equal Protection challenge. Protect Fayetteville v. City of Fayetteville, 510 S.W.3d 258 (Ark. 2017); 

see also infra note 162 and accompanying text. 

 
159 ARK. CODE ANN. §§ 14-1-401–403 (2015). 

 
160 Id. at § 14-1-403(a). 

 
161 Id. at § 14-1-402(a). 

 
162 See John Lyon, New Law Seeks to Bar Anti-Discrimination Ordinances, But Interpretations Vary, ARK. 

NEWS (Jul. 22, 2015), http://www.arkansasnews.com/news/arkansas/new-law-seeks-bar-anti-discrimination-

ordinances-interpretations-vary [https://perma.cc/2M6L-H3T6]. These cities argued that non-discrimination 

laws that protect LGBT individuals present no conflict with Act 137 because sexual orientation and gender 

identity are, in fact, protected classes in Arkansas state law. The state has an anti-bullying law that protects 

students from harassment based on sexual orientation and gender identity as well as a domestic violence law 

that prevents shelters from discriminating based on sexual identity. See ARK. CODE ANN. § 6-18-514 (2012); 

Domestic Peace Act, ARK. CODE ANN. § 9-4-106 (2012). This argument was taken up by the city of 

Fayetteville in Protect Fayetteville v. City of Fayetteville, No. CV2015001510, 2016 WL 9560283, at *2 

(Ark. Cir. Ct. Mar. 1, 2016), in defense of Ordinance 5781. The Arkansas Supreme Court was unconvinced 

by this claim and found the anti-discrimination laws to be in conflict with state law. Protect Fayetteville, 510 

S.W.3d at 263. 

 
163 No. CV2015001510, 2016 WL 9560283 (Ark. Cir. Ct. Mar. 1, 2016). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         237 

clause of the Fourteenth Amendment.164 The Circuit Court dismissed the constitutional 

question at the summary judgment stage, so the issue was not preserved for appeal to the 

Arkansas Supreme Court.165 Consequently, the Supreme Court remanded the 

constitutional question back to the circuit court where the case remains open.166 As 

discussed in Part II, it seems unlikely that the Circuit Court will find the law 

unconstitutional on Equal Protection grounds, given that the Act presents as facially 

neutral and makes no reference to LGBT individuals, or even to “sex,” unlike other 

similar legislation.  

 

Another angle is to use the Establishment Clause. This Part considers the likelihood 

of a successful Establishment Clause challenge to Act 137 under current jurisprudence 

and concludes that it is unlikely to prevail. It then charts a way forward, suggesting how 

the Establishment’s Clause legislative inquiry standard could be modified in order to 

ensure that the clause’s purpose is effected and to empower localities to prevent 

discriminatory preemption like that occurring in Arkansas. It then also suggests that a 

reconceptualization of our approach to state-local federalism could strike a balance 

between allowing cities to implement and experiment with local policy and upholding the 

principles of uniformity and state control on which preemption doctrine rests. 

 

A. Challenging Act 137 Under the Current Case Law  

 

It seems unlikely that the Supreme Court would find Act 137 unconstitutional. 

Despite the strength of the evidence of impermissible religious motivation, this Court 

would likely not invalidate the bill under the Lemon test, nor is it likely to apply one of its 

other Establishment Clause tests.  

 

1. The Secular Purpose Prong 

 

Demonstrating that Act 137 fails the first prong of the Lemon test—requiring that 

legislation have a legitimate secular purpose—is likely to be the most persuasive 

argument against its constitutionality. Relying on the same kind of evidence that the 

Court used in Edwards v. Aguillard,167 the facts surrounding the passage of Act 137 

strongly suggest that the Intrastate Commerce Improvement Act was not, in fact, 

 
164 Id. at *2. 

 
165 Id. 

 
166 Protect Fayetteville v. City of Fayetteville, 510 S.W.3d 258 (2017). 

 
167 107 S. Ct. 2573 (1987). 



 

238 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

substantially motivated by concern for the regulation of intrastate commerce, but instead 

by a desire to harm LGBT persons and protect religious prerogatives to discriminate. The 

bill’s legislative history and the statements of its sponsors support the conclusion that Act 

137 is a poorly disguised effort to allow discrimination against LGBT citizens on the 

basis of the religious beliefs of certain communities and legislators. 

 

During the bill’s debate on the Arkansas House floor, Representative Bob Ballinger, 

one of the bill’s primary sponsors, specifically framed the issue as one of LGBT rights. In 

his statement on the House floor, Representative Ballinger described supporters of the 

bill as representing “the will of the people” having to “fight” against money coming 

“from the Human Rights Campaign” (the nation’s largest LGBT advocacy group).168 He 

went on to discuss “the experience in Fayetteville” and acknowledged that he did not 

approve of “the certain direction we are moving in as a country.”169 

 

One of the bill’s other sponsors, Senator Bart Hester, explicitly told a legislative 

panel that he filed the bill in reaction to Ordinance 119’s protection of LGBT citizens 

because he felt that the bill was “anti-religious.”170 Senator Hester stated that the bill 

“isn’t about overreach, it’s about protecting the basic rights of the people in these 

municipalities, their basic rights of religious freedom.”171 Here, Senator Hester makes 

explicit that the bill was not motivated by a desire to regulate intrastate commerce, but by 

a concern for religious liberty. 

 

During the same House debate, others in support of the bill directly tied their support 

to their religious beliefs and current issues regarding religious liberty and discrimination 

against LGBT persons, such as by referencing Masterpiece Cakeshop.172 Representative 

 
168 State of Ark. House of Representatives, Feb. 13, 2015, WATCH LIVE (Feb. 13, 2015), at 11:22:00, 

https://sg001-harmony.sliq.net/00284/harmony/en/PowerBrowser/PowerBrowserV2/20160329/-1/1066#info_ 

[https://perma.cc/9UCH-95DR0]. 

 
169 Id. at 11:23:22. 

 
170 Bill to Ban Anti-Discrimination Laws at City, County Level Advances, ARK. NEWS (Feb. 5, 2015), 

http://www.arkansasnews.com/news/arkansas/bill-ban-anti-discrimination-laws-city-county-level-advances 

[https://perma.cc/UQH9-FKGE]. 

 
171 Id. 

 
172 Masterpiece Cakeshop, Ltd. v. Colo. Civil Rights Comm’n, 138 S. Ct. 1719 (2018). Masterpiece 

concerned a Colorado baker’s refusal to bake a wedding cake for a same-sex couple. The couple filed a 

complaint to the Colorado Civil Rights Commission, which concluded that the baker had violated Colorado’s 

civil rights law and fined the baker. The complaint was appealed and eventually the Supreme Court granted 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         239 

Copeland, speaking against the bill said, “I feel like if you were to ask the baker, if you 

were to ask the wedding planner . . . who have been attacked because of their religious 

beliefs, if they could stand here today, I feel like they would tell you that this is very 

much needed.”173 Echoing similar concerns for religious liberty, Representative Bentley, 

also speaking against the bill, after opening her remarks by stating that she is a 

conservative Christian, noted that: 

 

The reason we’re bringing this forth is because little businesses out 

there—a baker, or a pastor—a pastor that holds a firm conviction that the 

word of God says that gay marriage is wrong . . . I don’t think that pastor 

should have to form a marriage that goes against their strong religious 

convictions . . . [T]hose are the things that we’re talking about today . . . 

A baker that loves the word of God, that’s bringing her children up to 

honor God, and to worship God, should [not] have her business 

destroyed because she doesn’t want to bake a cake for somebody who’s a 

transgender trying to marry somebody else . . . I’m a Christian, this is the 

United States of America and it’s time that we stand up and say enough 

is enough.174  

 

Like Representative Bentley, Representative Copeland spoke directly to the purpose of 

the bill being to protect religious expression, not to regulate commercial interests. 

 

Of course, the statements of individual House members do not necessarily speak to 

the reasons that the entire body voted in favor of a piece of legislation. However, only 

four members of the House spoke on behalf of the bill during the floor debate, so the 

statements of Representatives Copeland and Bentley represent half of the on-the-record 

support for Act 137.175 Furthermore, many of the Representatives who spoke against the 

bill did so in terms of protecting LGBT rights, not within the framework of regulating 

 
certiorari. The Court did not decide on the merits of the case, which concerned whether a public 

accommodation law compelling a business owner to violate his sincerely held religious beliefs was a 

violation of the Free Speech and Free Exercise clauses of the Constitution, but instead remanded the case 

back to the lower court on procedural grounds. 

 
173 State of Ark. House of Representatives, Feb. 13, 2015, WATCH LIVE (Feb. 13, 2015), at 11:17:00, 

https://sg001-harmony.sliq.net/00284/harmony/en/PowerBrowser/PowerBrowserV2/20160329/-1/1066#info_ 

[https://perma.cc/9UCH-95DR0]. 

 
174 Id. 

 
175 See id. 



 

240 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

NDOs to ease the burden on business, further suggesting that both the purpose and 

impact of Act 137 was non-commercial. In fact, none of the testimony, either in favor or 

opposed to the bill, spoke to how Act 137 would affect intrastate commerce.  

 

Nonetheless, current precedent suggests that the statute would be upheld. In Cutter v. 

Wilkinson, the Court rejected an Establishment Clause challenge to the Religious Land 

Use and Institutionalized Persons Act (RLUIPA).176 Finding that RLUIPA, which 

prohibited the government from imposing substantial burdens on the ability of prisoners 

to practice their religions, was constitutionally permissible, the Court emphasized that the 

law did not grant privileges to religious persons, but merely eliminated government-

imposed burdens.177 Therefore, even if the Court agrees that Act 137 has a religiously 

motivated component, it may find that it is not religiously impermissible, given that it 

allows business owners the ability to discriminate against LGBT persons on the basis of 

their religious beliefs, but creates no affirmative requirements as to who they can or 

cannot serve or hire. It is possible that the Court would see the effects of Act 137 as 

analogous to those in RLUIPA—eliminating a government-imposed burden (from pro-

LGBT municipal ordinances) on the exercise of religious beliefs while not explicitly 

granting any privileges to religious persons. 

 

Moreover, the Court’s recent decision in Trump v. Hawaii178 suggests it may not be 

especially interested in looking behind the curtain of lawmaking motivation. In that case, 

the Court upheld President Trump’s Executive Order instituting a “travel ban” against 

nationals from seven predominantly Muslim countries.179 Surrounding the Executive 

Order (EO) were a series of statements made by the President, both during his 

presidential campaign and in the weeks leading up to the EO’s issue, that suggested that 

the ban was little more than pretext for discriminating against Muslim immigrants.180 

These statements played a large role in the plaintiff’s assertion that the official objective 

of the EO was to discriminate against Muslims and so violated the Establishment 

 
176 544 U.S. 709 (2005). 

 
177 Id. at 720. 

 
178 138 S. Ct. 2392 (2018). 

 
179 Id. 

 
180 See Jenna Johnson & Abigail Hauslohner, ‘I Think Islam Hates Us’: A Timeline of Trump’s Comments 

About Islam and Muslims, WASH. POST (May 20, 2017), https://www.washingtonpost.com/news/post-

politics/wp/2017/05/20/i-think-islam-hates-us-a-timeline-of-trumps-comments-about-islam-and-

muslims/?utm_term=.26dca928 [https://perma.cc/885V-6JWC]. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         241 

Clause.181 Employing rational basis review to analyze the EO, the Court held that the 

proclamation had an expressly legitimate purpose in protecting national security.182 

According to the Court, “because there is persuasive evidence that the entry suspension 

has a legitimate grounding in national security concerns, quite apart from any religious 

hostility, we must accept that independent justification,” the President’s inflammatory 

statements about Muslims notwithstanding.183 The decision in Trump v. Hawaii does not 

bode well for an Establishment Clause challenge against Act 137 based on legislative 

motivation under the Lemon test’s first prong. 

 

Granted, Hawaii is distinguishable on the facts from the circumstances surrounding 

Act 137 and does not necessarily bear directly on what the Court would do with a 

challenge to it. The Muslim Ban case was complicated by the fact that it involved 

national security and immigration powers, issues on which the Court generally defers to 

the executive.184 Act 137 does not present issues tied to executive powers. However, 

Hawaii is the most recent Establishment Clause case in which the Court has had both 

opportunity and grounds to probe lawmaking motive and it declined to do so. The 

willingness of the Court in Hawaii to put aside the President’s “charged statements” 

about Muslims may be indicative of a broader reluctance to cast aside proffered 

objectives in favor of an independent inquiry in motivation, particularly in the 

Establishment Clause context.185 

 

2. The Principal or Primary Purpose Prong 

 

Act 137 would also likely pass constitutional muster under the second prong of the 

Lemon test. Act 137 does not resemble the types of laws that have historically been 

invalidated under this prong. Because Act 137 impacts all citizens equally in that it 

doesn’t target specific groups but merely requires that all municipalities refrain from 

creating any protections, including but not limited to protections for LGBT individuals, 

that do not exist in state law, it is unlikely that a court would find that the primary 

purpose of the bill is to advance a particular religion. 

 

 
181 Hawaii, 138 S. Ct. at 2406, 2417. 

 
182 Id. at 2421. 

 
183 Id. 

 
184 Id. at 2408. 

 
185 Id. at 2447 (J. Sotomayor, dissenting). 



 

242 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

In an article titled “Challenging and Preventing Policies that Prohibit Local Civil 

Rights Protections for Lesbian, Gay, Bisexual, Transgender and Queer People,” Professor 

Jennifer L. Pomeranz argues that a challenge could be raised under this prong on the 

theory that “Act 137 imposes economic and other burdens on LGBTQ persons and their 

families to advance the religious beliefs of those accommodated. Through Act 137, the 

state has given the force of law to religious business owners who wish to discriminate 

against same-sex couples.”186 She relies on both Estate of Thornton v. Caldor187 and 

Corp. of Presiding Bishop of Church of Jesus Christ of Latter-day Saints v. Amos188 for 

the proposition that “the state cannot give the force of law to a person’s religious belief 

and require accommodation by others regardless of the burden.”189 This principle does 

emerge from the two cases, but the facts of each suggest that the Court would not find 

that the effect of Act 137 gives religious business owners the force of law. 

 

In Caldor, the Court struck down a Connecticut law that provided that no individual 

had to work on the day of the week which he considered his Sabbath.190 The Court held 

that this violated the second prong of the Lemon test and so was unconstitutional under 

the Establishment Clause.191 The Court’s reasoning emphasized that the statute imposed 

an affirmative and absolute duty on business owners to conform their business practices 

to the religious observances of their employees.192 Because the statute placed the interests 

of Sabbath observers over all other interests, the Court concluded that the law 

impermissibly advanced a particular religious practice.193  

 

In Amos, the plaintiffs alleged that Section 702 of the Civil Rights Act of 1964 

violated the Establishment Clause by allowing religious employers to consider religion in 

 
186 Jennifer L. Pomeranz, Challenging and Preventing Policies That Prohibit Local Civil Rights Protections 

for Lesbian, Gay, Bisexual, Transgender, and Queer People, 108 AM. J. PUB. HEALTH POL. 67, 69 (2018). 

 
187 472 U.S. 703 (1985). 

 
188 483 U.S. 327 (1987). 

 
189 Pomeranz, supra note 186, at 69. 

 
190 Estate of Thornton v. Caldor, 472 U.S. 703, 710–11 (1985). 

 
191 Id. at 709. 

 
192 Id. 

 
193 Id. 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         243 

choosing employees for non-religious jobs.194 The Court held that this was not an 

Establishment Clause violation.195 The Court specifically distinguished the case from 

Caldor, writing in a footnote that the obligation imposed on business owners by the 

statute at issue in Caldor did not exist in the Civil Rights Act.196 Because the statute 

imposed no duty, the mere fact that the law allowed churches to exercise discretion in 

advancing religion was not enough to violate the Establishment Clause.197 

 

Act 137 looks a great deal like the statute at issue in Amos. Unlike in Caldor, Act 137 

does not impose any kind of affirmative duty or obligation on business owners in 

Arkansas. It does not require that business owners discriminate against LGBT individuals 

or turn away customers due to their religious beliefs. It does require, however, that LGBT 

employees or customers of these businesses accept the religious observances of their 

bosses and business proprietors, or risk being fired or denied services. Both the statute in 

Amos and Act 137 essentially give license to business owners to discriminate according 

to their religious beliefs. As such, given its holding on similar facts in Amos, the Court 

would likely not find reason to strike down Act 137 under the Lemon test’s second prong. 

 

a. The Endorsement Test  

 

The Court could, either in place of or as a supplement to the first two prongs of the 

Lemon test, choose to analyze Act 137 using O’Connor’s endorsement test. Using the 

“reasonable observer” standard that the endorsement test establishes, there is a 

compelling argument to be made that an individual with knowledge of not only the 

specific act at issue, but also of the general history and climate surrounding the 

challenged act, could interpret the Arkansas legislature’s actions as an endorsement of a 

particular brand of Christianity.198 However, given that the Court applies the endorsement 

 
194 Corp. of Presiding Bishop of Church of Jesus Christ of Latter-Day Saints v. Amos, 483 U.S. 327, 331 

(1987). 

 
195 Id. at 339–40 

 
196 Id. at 337 n.15. 

 
197 Id. at 337. 

 
198 In her concurrence in Capital Square Review & Advisory Board v. Pinette, a case involving a cross 

displayed on public land by the Ku Klux Klan, Justice O’Connor further clarified the “reasonable observer” 

standard, writing that “the reasonable observer in the endorsement inquiry must be deemed aware of the 

history and context of the community and forum in which the religious display appears.” 515 U.S. 753, 780 

(1995) (O’Connor, J., concurring). 



 

244 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

test infrequently and that current trends indicate that it is falling out of favor,199 the 

likelihood of the Court adopting an endorsement test framework to assess the 

constitutionality of Act 137 is low. 

 

3. The Excessive Entanglement Prong  

 

The excessive entanglement prong would likewise probably not provide a basis on 

which to find Act 137 unconstitutional. Act 137 creates no ongoing monitoring or 

enforcement program between the state and a religious organization or program, and so 

does not implicate the administerial concern of the excessive entanglement prong. Nor 

does it violate the political divisiveness standard. Pomeranz argues that Act 137 

implicates the type of political entanglement that the Establishment Clause was designed 

to protect against because of the religious character of many of the statements made on 

the House floor during the bill’s debate.200 However, given that the Court has never 

invalidated a law on the political divisiveness standard alone, and given that the times it 

has invoked the doctrine in its reasoning have involved fact patterns with much closer 

association between the political and the religious, it is unlikely that Act 137 would be 

found impermissible for reasons of creating an excessive government entanglement with 

religion. Ultimately, under the Supreme Court’s current Establishment Clause 

jurisprudence, Act 137 would likely be upheld under all three prongs of the Lemon test. 

In the absence of a modified standard of legislative inquiry in Establishment Clause 

cases, such as that suggested below, Act 137 is constitutional. 

 

B. A Modified Approach to Both Legislative Motive and Preemption  

 

1. Moving Toward a But-For Inquiry for Legislative Motivation 

 

To return to Judge Parr’s observation from the Introduction, courts need not be struck 

with blindness when confronting patently discriminatory legislation.201 The reality is that 

overtly religious legislation is extremely rare, and most legislators are knowledgeable 

enough to recognize the need for pretextual justifications for discriminatory laws. This 

does not mean, however, that we must be resigned to the fact of their validity. Courts 

should not be blind to religious bias, which is evident, even if that evidence is occluded 

by assurances of secularism. Furthermore, especially in cases like Act 137 where 

 
199 See supra Part II.C.iv. 

 
200 Pomeranz, supra note 186, at 70. 

 
201 See Parr v. Mun. Ct., 479 P.2d 353, 356 (1971). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         245 

legislators made no attempts to hide their religious motivations in the course of public 

debate, courts should employ an Establishment Clause standard that substantially 

interrogates legislative motive.  

 

A number of scholars have proposed modifications to the Court’s current 

Establishment Clause tests with varying degrees of expansiveness, plausibility, and 

administrability.202 One proposal has been that the Supreme Court incorporate into its 

Establishment Clause jurisprudence an inquiry for discriminatory intent similar to that it 

uses in equal protection cases. The equal protection framework for assessing legislative 

motivation essentially establishes a but-for test for legislative decision-making. In other 

words, government action is rooted in a constitutionally impermissible purpose if the law 

would not have been passed but-for the role that impermissible purpose played in the 

decision-making process.203 The specifics of such an inquiry are fact- and context-

specific.204 

 

In his article “Religious Purpose, Inerrancy, and the Establishment Clause,” Professor 

Daniel Conkle argues that this standard should migrate to the Establishment Clause, 

particularly for cases where a government action may have been influenced by multiple 

motivations.205 Act 137 is precisely such a case. Multiple motivations exist “when 

individual legislators act for permissible as well as impermissible reasons, and also when 

some legislators act for permissible reasons, but others for impermissible.”206 For Act 

 
202 See, e.g., Kent Greenawalt, Religiously Based Judgments and Discourse in Political Life, 22 J. CIV. 

RIGHTS & ECON. DEV. 445 (2007); Caroline Mala Corbin, Intentional Discrimination in Establishment Clause 

Jurisprudence, 67 ALA. L. REV. 299 (2015); Daniel O. Conkle, Religious Purpose, Inerrancy, and the 

Establishment Clause, 67 IND. L.J. 1 (1991). 

 
203 See Conkle, supra note 202, at 1. 

 
204 See, e.g., Vill. Arlington Heights v. Metro. Hous. Dev. Corp., 429 U.S. 252 (1977) (holding that the denial 

of a rezoning petition was possibly motivated by racial discrimination and that the case should be remanded 

for further consideration because the plaintiffs failed to carry their burden of establishing discriminatory 

intent). The court noted that: 

 

Rarely can it be said that a legislature or administrative body operating under a broad 

mandate made a decision motivated solely by a single concern, or even that a particular 

purpose was the “dominant” or “primary” one . . . but . . . [w]hen there is a proof that a 

discriminatory purpose has been a motivating factor in the decision, this judicial 

deference is no longer justified. Id. at 265. 

 
205 Conkle, supra note 202, at 1. 

 
206 Id. 



 

246 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

137, for example, there are two discernible motives: the first is the permissible secular 

purpose of regulating intrastate commerce, and the second the impermissible desire of 

some legislators to limit the rights of LGBT persons on the basis of their religious beliefs. 

Applying the equal protection test to Establishment Clause cases would enable courts to 

both continue to show deference to legislatively proffered purposes—if the facts suggest 

they are sincere and there is an absence of evidence to suggest impermissible motives—

and to look behind the legislative curtain where such evidence does exist. 

 

Of course, the kind of counterfactual inquiry required by a but-for test necessitates 

speculation. The inherent difficulty of speculative tests could potentially be lessened here 

by establishing some sort of trigger for when a more searching inquiry into legislative 

motive is warranted. For example, the presence of religious statements on the legislative 

floor or from legislators during the promotion of the bill could be used to justify a deeper 

inquiry into an asserted rationale. These triggers would not act as evidence in and of 

themselves for Establishment Clause violations—although they could be used in making 

that determination—but would merely provide a basis on which a court could turn to the 

but-for test. 

 

In the case of Act 137, it is not a stretch to imagine that in the absence of religiously 

informed beliefs regarding LGBT individuals, the Arkansas legislature would not have 

addressed Fayetteville’s NDO with such urgency, or at all. Furthermore, even if we 

accept the Arkansas legislature’s stated purpose of standardizing non-discrimination 

protections statewide, in a universe where religious beliefs played no role in decision-

making, incorporating sexual orientation and gender identity protections into state law 

would have been an equally viable method of achieving uniformity as preempting 

Fayetteville’s ordinance. There is no evidence that the Arkansas legislature ever 

considered this option. The jump to preemptive action when other appropriate means 

existed suggests that the legislature was not interested in standardization for 

standardization’s sake, but standardization as a means by which to legitimize 

discrimination.  

 

2. Rethinking Preemption and Local Legislative Authority 

 

The controversies surrounding Arkansas’ Act 137, the Equal Access to Intrastate 

Commerce Act in Tennessee, and North Carolina’s HB2 are only three examples of a 

tension that continues to build between state and municipal governments. The path to 

successful repeal on federal constitutional grounds for these bills is, as has been 

demonstrated, extremely narrow. Nonetheless, these laws flout both principles of local 

autonomy and of non-establishment, ideals that have animated the American ethos since 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         247 

the nation’s founding. As it stands, our current standards provide little protection against 

either preemptive or obviously pretextual legislation. Our approaches to both state-local 

preemption and to Establishment Clause inquiries into religiously motivated lawmaking 

should be adapted to account for modern political realities, including profound political 

polarization and an Executive branch that enables and empowers certain religious 

adherents over others. 

 

3. Protecting the Autonomy of Municipalities 

 

During the House debate over Act 137, Representative David Whitaker took the floor 

to speak in opposition to the bill’s passage. He urged the body to vote against the Act 

because the ordinance it sought to preempt had been produced by a local democratic 

process that deserved to be respected, saying, “I do have an obligation as one of the 

representatives of the constituency and voters of Fayetteville to express to you their 

desires to run their own affairs.”207 He continued: “My constituents beg you: let them run 

their own affairs and we would definitely love to extend that courtesy to you, that your 

cities, your counties, your school districts, heck, your neighborhood associations, can 

govern yourselves better than we can.”208 Representative Whitaker’s statement raises 

fundamental issues regarding the relationship between state and local governments and 

presents one example of the kind of city/state conflicts that are playing out across the 

country.  

 

In his article, “The Challenge of the New Preemption,” Professor Richard Briffault 

argues that the aggressive preemption we have seen from states in the past decade 

requires a reconsideration of how our system protects local governments and self-

determination, especially in a time of increasing ideological polarization.209 Professor 

Briffault writes that a system that provides little protection for municipal lawmaking 

authority, which is what most states have, may work when municipal and state 

governments are generally cooperative, but “when legislatures seem fraught with open 

hostility in a way they haven’t been in the past, the traditional laissez faire approach risks 

jeopardizing the ability of local governments to play their key role in our system.”210  

 
207 State of Ark. House of Representatives, Feb. 13, 2015, WATCH LIVE (Feb. 13, 2015), at 11:06:00, 

https://sg001-harmony.sliq.net/00284/harmony/en/PowerBrowser/PowerBrowserV2/20160329/-1/1066#info_ 

[https://perma.cc/9UCH-95DR0]. 

 
208 Id. 

 
209 Richard Briffault, The Challenge of the New Preemption, 70 STAN. L. REV. 1995 (2018). 

 
210 Id. at 2017. 



 

248 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

Even in the early twentieth century political thinkers lamented the role of cities in 

their own legislative affairs. In 1912, political scientist Charles Beard wrote “not only do 

state legislatures interfere with the fundamental rights and pettiest details of city affairs, 

but their consent is required for some of the most insignificant undertakings of municipal 

government.”211 The ensuing century has in many ways exacerbated Beard’s complaints. 

Professor Kenneth Stahl sums up the current difficulty in the American city/state 

relationship: “Today, with Democrats controlling most cities and rural Republicans 

controlling most states, the traditional rivalry between urban and rural areas in American 

politics is now expressed vertically in the relationship between the state and its cities. 

That is evident in the preemption battles.”212 Especially as states have begun to move 

beyond traditional preemption laws to enacting punitive213 and nuclear214 preemptive 

legislation, it is clear that something has to give in the current state/local dynamic. 

 

One imperfect solution would be an expanded understanding of home rule authority 

for municipalities to encourage greater protection of local lawmaking authority and allow 

the legal status of local governments to be “bolstered.”215 The boundaries of home rule 

should be determined with an eye, first and foremost, toward “the empowered local self-

government.”216 This would create limits on state preemption while respecting the state’s 

ultimate authority to set statewide policy.217 Some state governments and high courts 

have already proved themselves amenable to increasing protections for local lawmaking 

authority. In California, the state Supreme Court ruled that preemption is limited to 

situations where the legislated subject is one of “statewide concern” and the statute that 

addresses it is “reasonably related to its resolution.”218  

 
211 Alan Ehrenhalt, What Do States Have Against Cities, Anyway?, GOVERNING (Nov. 2017), 

www.governing.com/columns/assessments/gov-urban-penalty-legislatures-cities.html 

[https://perma.cc/4PMJ-2HVZ]. 

 
212 Stahl, supra note 23, at 144. 

 
213 Punitive preemption not only negates local laws but also punishes governments or local officials for 

enacting the preempted policies in the first place. Briffault, supra note 209, at 2002. 

 
214 Nuclear preemption refers to preemptive legislation that completely eliminates local lawmaking authority, 

either over particular fields or with respect to certain subject matter. Id. at 2007. 

 
215 See id. at 2018. 

 
216 Id. 

 
217 See id. 

 
218 See Cal. Fed. Sav. & Loan Ass’n v. City of L.A., 812 P.2d 916, 925 (1991). 



 

39.1 COLUMBIA JOURNAL OF GENDER AND LAW                         249 

The Ohio Supreme Court took an even more limiting approach to state preemption, 

holding that the state legislature cannot preempt local legislation unless the state law is of 

a “general nature,” which the Court interpreted to mean “statutes setting forth police, 

sanitary or similar regulations and not statutes which purport only to grant or to limit the 

legislative powers of a municipal corporation to adopt or enforce police, sanitary, or other 

similar regulations.”219 On these grounds, the Ohio Court struck down state attempts to 

overturn local regulations concerning nutrition, business regulations, and hiring 

requirements.220 The California and Ohio approaches, though they have not been widely 

adopted, would be a step toward reconciling the need for state governments to legislate 

uniformly with the function that local governments serve in our system as policy 

innovators with local autonomy. 

 

Granted, empowering municipalities is neither a panacea nor entirely positive. There 

are problems associated with increased local control, such as non-uniformity of laws 

within the state, which can place heavy burdens on regulated parties, the risk of 

parochialism, and the seeming arbitrariness of municipal borders, where the regulations 

of one municipality can impact its neighbors.221 It is reasonable to be wary of increased 

municipal power. However, and particularly in the context of the aggressive preemption 

action that has been taken by states in recent years, the benefits accrued by granting 

municipalities more flexibility in legislating are greater than the risks incurred. Cities, 

with the consent of support of their citizens, should be allowed to experiment with 

legislative solutions to local problems, respond to local concerns, and protect their 

vulnerable populations.  

 

Of course, even a modified and expanded preemption standard might not ensure that 

progressive ordinances like Fayetteville’s go un-preempted by state governments, but it 

would create greater latitude in the state-local system for cities to experiment with policy 

at the local level and to serve the needs and interests of their particular communities with 

a lessened fear of interference from the state. Furthermore, limiting the scope of 

legitimate state preemption would serve many underlying values of our federalist system, 

such as encouraging citizen participation in democracy and enabling responsive 

government.222 

 
219 See Canton v. State, 766 N.E.2d 963, 968 (2002). 

 
220 See Briffault, supra note 209, at 2013. 

 
221 See Richard Briffault, Our Localism, Part II: Localism and Legal Theory, 90 COLUM. L. REV. 346, 446 

(1990). 

 
222 See generally Briffault, supra note 209. 



 

250 COLUMBIA JOURNAL OF GENDER AND LAW 39.1 

 

CONCLUSION 

 

A number of scholars who work in this area suggest that, when the right facts come 

along, an Establishment Clause challenge should be considered for these kinds of non-

discrimination preemption laws,223 but it is hard to imagine a more damning set of facts 

than those presented by Act 137. Legislators generally understand that legislating solely 

on religious grounds is unconstitutional and adapt accordingly, so if advocates wait for 

cases that present more evidence of religious motivation than that surrounding Act 137 to 

explore the merits of an Establishment Clause challenge, they may wait forever. The 

flaws of current Establishment Clause jurisprudence and the landscape of the current use 

of preemption beg for a more radical judicial approach to parsing prejudice from pretext 

when assessing legislative motivation, especially when it comes to religiously motivated 

laws. Federal courts have been historically reluctant to interfere in assessing the proper 

balance of state and municipal power, deferring to principles of federalism in leaving 

those determinations almost entirely up to states. Nonetheless, courts are well within their 

bounds to look behind the curtain, so to speak, of legislative motivation when such an 

inquiry is necessary. 

 

Allowing the but-for legislative inquiry test to migrate from Equal Protection Clause 

to Establishment Clause jurisprudence would give the Establishment Clause a great deal 

more efficacy as an avenue for striking down discriminatory state and local law. This 

more searching legislative motivation inquiry in combination with a broadened approach 

to state-local relations would help ensure that, through regulation of things like local non-

discrimination laws without fear of preemption—and wider avenues for judicial 

challenge if they are preempted—cities are able to effect community values and protect 

their most vulnerable populations. 

 
223 See BRIFFAULT ET. AL, supra note 4, at 13; Legal Strategies to Counter State Preemption and Protect 

Progressive Localism: A Summary of the Findings of the Legal Effort to Address Preemption (LEAP) Project, 

A BETTER BALANCE (Aug. 9, 2017), https://www.abetterbalance.org/resources/legal-strategies-to-counter-

state-preemption-and-protect-progressive-localism-a-summary-of-the-findings-of-the-legal-effort-to-address-

preemption-leap-project/ [https://perma.cc/YMV2-TPQ3].