Paul A. Diller, The Political Process of Preemption, 54 U. Rich. L. Rev. 343 (2020).

Click here to download PDF

Paul A. Diller *


Preemption, particularly of the state-city variety, has become a hot topic. State legislatures in many states over the last decade have preempted a wide swath of areas in which cities and counties were previously free to govern. In addition to the sweeping nature and frequency of preemption, the increasingly aggressive methods of enforcing preemption have drawn notice. The threat of fiscal penalties, removal of local officials from office, and even criminal sanctions constitute what one scholar has dubbed the phenomenon of “hyper preemption.”[1]

There are several reasons why this new landscape of preemption is of concern. Advocates of local control lament the loss of cities’ ability to use their local expertise to solve local problems. Relatedly, fans of local innovation fear that aggressive preemption limits the ability of cities to try out new policy proposals that might work their way up the state or federal ladder. The specter of personal penalties for city councilors who support preempted ordinances may dissuade citizens from running for local elected office, a crucial training ground for state and federal positions and an intrinsically important component of our democratic system. Even if not punitive, sweeping substantive preemption may weaken cities so much that voters and potential elected officials have less interest in participating in city governance. Finally, due to the political valence of much of the recent preemption—by more conservative state legislatures preempting the policy priorities of populous, politically liberal cities—political progressives see preemption as an attack on their values and priorities.[2]

Because of the issue-based nature of preemption, it is tempting to approach the subject in a results-based manner. Political progressives, for instance, may oppose preemption of local firearm regulation, but cheer preemption of local ordinances that crack down on illegal immigration. Conversely, political conservatives bemoan preemption of local “right-to-work” ordinances, but support preemption of the minimum wage.[3] Courts and legal scholars have struggled for years to offer a neutral way out of this morass.

For decades some scholars and courts have argued that a neutral way of determining the legitimacy of statewide preemption is by focusing on whether the particular issue in play is “local” or not.[4] Many scholars and judges, however, have rightly criticized this approach as intellectually unsatisfying.[5] Another view, recently offered in a brilliant essay by Nestor Davidson, is that courts might find their way out of the results-based morass by homing in on whether the preemptive legislation furthers the “general welfare” of the state.[6] This approach is attractive, particularly with respect to deregulatory preemption.[7] Employing it may still, however, require courts to make policy-based determinations regarding whether a state or local choice furthers the general welfare.[8]

This Article suggests a different path out of the morass by inviting scrutiny of the political process of preemption. Focusing on state rather than federal preemption, this Article assumes—without endorsing—the primacy of the state over its political subdivisions, a view embraced by the Supreme Court in the seminal 1907 case of Hunter v. City of Pittsburgh and beyond.[9] While making this assumption, this Article nonetheless critically examines it by digging deeper into the roots of state sovereignty and the state’s claim to dominion over its subentities.

At the time of the Founding, states were seen as sovereign due to their representation of “the people.” Building off of this conception of state sovereignty, the Court in Hunter recognized plenary state power over cities—or local “agencies”—as the Court referred to them.[10] Fifty-seven years later, in Reynolds v. Sims, the Court recognized one-person, one-vote as a key component of representation, embracing the idea that the state legislature ought to be responsive to the policy preferences of a majority of the state’s voters.[11] After Reynolds, the Court continued to reaffirm Hunter’s notion of state primacy without explicitly re-examining the premises of this assumption. After Reynolds, however, state primacy and the majoritarianism of state government had become fused, and rightly so. Only a democratically legitimate state government—that is, one which purported to represent credibly a majority of the state’s population—could justifiably exercise its plenary powers over the democratic subunits within it.

Working from these premises, this Article argues that when state preemption is the product of a credibly majoritarian lawmaking process, it might be considered less objectionable, regardless of its political valence. A key component of democratic illegitimacy, this Article will argue, is intentional political gerrymandering. Such gerrymandering distorts the legislature’s composition and leads to legislative products that do not accurately reflect the views of the median voter statewide. When produced by other processes, such as direct democracy, preemption may reflect the majority’s will in some crude sense, but may raise other normative concerns.

Before proceeding, it is useful to clarify what this Article means by preemption. This Article uses the word broadly to include any override of pre-existing local power or prerogative by statute or constitutional amendment. Such preemption may impose a new regulatory regime from above, displacing the locality’s previously governing regime, or it may impose a regulatory regime when none such existed previously at the local level. As used here, preemption may also simply deprive the locality of the authority to implement a regulatory regime or fiscal choice, such as its preferred level of taxation, without providing any new regime or supplementary revenue in its place.[12] Almost all of the preemption discussed here is of the general type—i.e., not special legislation that formally applies only to one city or a handful of cities—but the framework of the analysis should apply similarly to special preemption.[13]

Part I of this Article examines the roots of the state’s role as sovereign, and how the democratic legitimacy of the state legislature eventually became linked to a notion of majoritarianism, as epitomized by Reynolds. Part II discusses how and why gerrymandering and political geography may cause a legislature to stray from majoritarianism. Part II examines this phenomenon in the context of state-local relations, in particular. Part II includes vignettes of some key states where gerrymandering has arguably led to unrepresentative preemption in approximately the last decade. Part III looks at how the initiative process differs from the legislative process in the context of preemption. Nothing is more seemingly majoritarian than the voters enacting a policy preference by a majority vote. Nonetheless, direct democracy raises distinct problems of democratic legitimacy, which Part III assesses in the context of state-local relations. Finally, Part IV looks at some other factors in prominent preemption instances of late, including legislative “horse trading” and punitive enforcement mechanisms, and asks what these dynamics reveal about the legitimacy of state overrides.

I. State as (Popular) Sovereign

In American constitutional theory, it was the states, rather than the federal government, that traditionally functioned as indestructible sovereigns. The states formed the Union, it was thought, and thus, they might also abolish it or at least voluntarily secede from it.[14] While the Civil War established the Union’s primacy as a sovereign when push came to shove, states retained their roles as residual sovereigns, endowed with attributes of sovereign-like immunity and, in contrast to the federal government, virtually unlimited default lawmaking authority.[15]

With respect to states’ lawmaking authority, after the Revolution, states were seen as having inherited the plenary powers enjoyed by the British Parliament.[16] Of course, at the time of the Revolution, Parliament still operated in the shadow of a monarch, and thus had to pay at least lip service to the idea that it had been invested with power by the head of state. This king or queen, in turn, had historically laid claim to divine right as the source of his or her authority to rule.[17] During a centuries-long process that accelerated during the English Civil War, the dominant narrative supporting the legitimacy of English government fitfully shifted from the monarch serving as God’s agent on earth to an elected Parliament that derived its power from “the people.”[18] In the new United States, in the absence of a monarch, state legislatures, and eventually the national government, wholly embraced the notion of “the people,” or “popular sovereignty,” as the source of their governing legitimacy.[19]

A. Popular Sovereignty

It was easy enough to declare that “the people” ruled in the new United States. It was more difficult to define exactly who “the people” were and how the new government represented them. “The people” who participated in what passed for democracy at the Founding were, of course, a very small subset of all of the human beings residing in the young nation. While franchise qualifications varied by state (and sometimes city), those entitled to participate in democracy were generally white, male, and in possession of a minimum amount of real or personal property.[20] This meant that approximately six percent of Americans could vote during George Washington’s presidency.[21]

Hence, the fiction of popular sovereignty at the beginning of the nation’s history was sustained by a very narrow view of who counted as “the people.” Alternatively, even if one viewed all persons, or at least all free persons (given the perception of enslaved persons as property), as constituting “the people,” the smaller subset was entitled to act on their behalf.

In addition to a narrow understanding of which persons were allowed to participate in democracy, states at the Founding had to devise a way for “the people” to be represented in an assembly, or legislature, imbued with the power to make law for all. Unsurprisingly, the colonists had looked to the system whence they came—the British Parliament—as a model for representation, although they did not hesitate to improve upon it.[22] In Britain, the well-established practice was to use geographical units—rather than population—as the primary basis for selecting representatives, even if this system, with its over-represented “rotten boroughs,” had devolved into a “travesty” of representation.[23] In the colonies, elections were held more frequently, and “the assignment of representation kept pace with movement into new areas much more closely than in England.”[24] After the Revolution, most states continued the practice of using geography as the basis for representation, affording towns or counties equal, or at least minimum, representation regardless of population in at least one house of the state legislature.[25]

The Levellers during the English Civil War were the first to suggest that representation be based on population rather than geographical unit.[26] The method chosen for allocating seats in the U.S. House of Representatives at the Constitutional Convention, of course, largely embraced this approach.[27] Over the course of the next century and a half, more states adopted population as the basis for allocating seats in their state legislative chambers.[28] By 1920, state constitutions required that seventy-five percent of all state legislative chambers be allocated on the basis of population.[29] The use of geography to allocate legislative seats remained more common for the upper houses in states.[30] Despite the increasing commitment on paper to allocating seats on the basis of population, many states did not comply with their own state constitutions.[31] In Alabama, for instance, the state legislature did not reapportion between the 1900 and 1960 censuses.[32] Hence, while the United States population had become much more urbanized by the mid-twentieth century, state legislatures often did not reflect these changing demographics.[33]

Urban-centered reform groups led a sustained, multifront effort to correct the imbalance between the country’s population and its elected representatives throughout the twentieth century.[34] Their efforts culminated in the landmark victories before the United States Supreme Court in Baker v. Carr (1962), which held that apportionment was not a political question,[35] and Reynolds v. Sims (1964), which held that the Equal Protection Clause of the Fourteenth Amendment required state legislatures to comport with one-person, one-vote.[36] Also in 1964, the Court delivered another victory to urban voters when it interpreted Article I, section 2, to require that all congressional districts within a state be equipopulous.[37] Previously, many states had urban Congressional districts that were significantly more populous than those in rural areas.

While Reynolds settled on equal population as the appropriate method for electing state legislatures, it was, in retrospect, surprisingly vague about exactly which population mattered for these purposes. From the beginning of the Republic, states had adopted different approaches to the question of who the “people” were for the purposes of apportionment. Consistent with their sometimes narrow views of “the people,” several states in the early nineteenth century apportioned on the basis of “free white males” or “free white males over 21.”[38] A handful of states used taxes paid, rather than any subset of people, as the metric for apportioning seats.[39] During Reconstruction, many state constitutions used “qualified voters” as a basis for apportioning, but by the end of the nineteenth century, most states simply used total population for this purpose.[40]

Of course, while many states in the late nineteenth century expanded the basis for apportionment to include the entire population, they were at the same time excluding large numbers of persons from voting, such as blacks (particularly in the South) and women. As the franchise gradually extended to these previously excluded groups, such as women with the ratification of the Nineteenth Amendment in 1920,[41] and blacks in the South with the passage of the Voting Rights Act in 1965,[42] total population became a somewhat more accurate reflection of the population that could vote for a representative. To be sure, even now, there are many in the total population who are excluded from voting, such as minors, noncitizens, and certain convicted criminals in some states. Nonetheless, all states currently use total population to allocate seats, although there is some variation on the margins with respect to prisoners and temporary residents like nonresident military personnel and college students.[43]

In 2016, the Supreme Court rebuffed an attempt by litigants to force a state to use citizens-of-voting-age-population (“CVAP”), rather than total population, as the metric for apportionment, holding that this was not required by Reynolds.[44] The Court did not, however, require states to use total population as the metric for apportionment, thus leaving at least the theoretical door open for a state to use CVAP as the basis for apportionment. Legislators in several states, anticipating that the 2020 census might include a citizenship question, proposed that CVAP or something similar be used to apportion state legislative and Congressional seats after the census.[45] After an unfavorable Supreme Court decision,[46] President Donald Trump abandoned his earlier efforts to direct the Census Bureau to include a citizenship question on the 2020 census.[47]  Not completely deterred, however, Trump asked the Bureau to collect citizenship data from other sources in other ways so that interested states might use it for post-2020 redistricting.[48]

B. State as Sovereign over Localities

On a different doctrinal front, the state’s primacy over local governments would be solidified between the Founding and the early twentieth century. The rise of Dillon’s Rule and the Supreme Court’s embrace of the state’s hegemonic control over local governments in Hunter largely snuffed out the views of some prominent jurists, such as Judge Thomas Cooley, who had argued that there were constitutional limits on the state’s primacy over its cities.[49] Per Hunter, a state enjoys wide discretion to reorganize its local governments as it sees fit, whether consolidating them, disincorporating them, or redrawing their boundaries.[50] A state may even abolish local government entirely if it so chooses.[51]

After Reynolds v. Sims, state sovereignty over local governments, per Hunter, and popular sovereignty could be merged conceptually. Indeed, in rejecting Alabama’s and other states’ use of political subdivisions as the basis for representation, as opposed to equal population, Reynolds cited Hunter for the proposition that local entities are mere agents of the state.[52] Hence, unlike the sovereign states in our federal order, local governments did not possess sovereignty and thus did not need representation in their own capacity at the state level.[53] The flip side of this articulation is that the state’s supremacy is legitimate because its government is representative of a statewide majority. The state’s responsibility, therefore, for ruling its localities hinges on a credible claim to representing a majority of the state’s voters.

As the Supreme Court recognized in Avery v. Midland County, in 1968, state legislatures “exercise extensive power over their constituents and over the various units of local government.”[54] Although Avery recognized that there was a sphere of authority in which local governments primarily operated,[55] an implicit premise of the legitimacy of state legislatures’ power over their subdivisions was compliance with one-person, one-vote and the majoritarianism envisioned by Reynolds. Indeed, the dissenters in many of the cases that built upon Reynolds by extending one-person, one-vote to elections for local bodies like county commissions and community college boards invoked one-person, one-vote at the state level as a justification for state-sanctioned deviations from one-person, one-vote locally.[56] Later, in the key one-person, one-vote case of Ball v. James in 1981, Justice Powell provided the crucial fifth vote to uphold the constitutionality of the acreage-based voting scheme for the Phoenix-area’s Salt River District. He based his concurrence on his “expect[ation] that a legislature elected on the rule of one person, one vote will be vigilant to prevent undue concentration of power in the hands of undemocratic bodies,” and thus could be trusted to oversee the district.[57]

Of course, the proposition that majoritarianism legitimizes control of subentities is contingent upon the will of the majority being in some way discernible, itself a contested proposition among political scientists.[58] Working from the general assumption that a majoritarian will is knowable, Hunter allows the majority of the larger entity to trump the majority of the smaller entity. Through their state constitutions, some states have chosen to limit the reach of the state’s “Hunter power” by cordoning off certain areas that the state may not preempt, or imposing procedural restrictions on the state’s ability to control its local governments.[59] Many states, however, have no such constraints, and even in those that do, the legislature may still preempt local governments with respect to the vast majority of potential legislative subject matters.[60]

The Reynolds Court anticipated that by reapportioning seats from sparsely populated rural areas to more populous urban areas, the majority could speak more clearly.[61] Prior to Reynolds, at least in many states, rural voters exercised an outsized voice due to states’ allocation of legislative seats on the basis of local government units, or because state legislatures did not fulfill their responsibilities under their state constitutions to reapportion on the basis of population changes. There is limited evidence that Reynolds’s reallocation of political power succeeded in redistributing state resources from the formerly overrepresented areas to those previously underrepresented.[62] As discussed next, however, the Reynolds Court woefully underestimated the extent to which the process of districting itself could muzzle the voice of the “majority” that Reynolds sought to reveal.

II. How the Reynolds “Majority” Became Muzzled

An extensive political science literature has demonstrated that a winner-take-all, first-past-the-post system of district elections, like that used to elect the vast majority of state legislatures, can lead to a very different legislative composition from that which other systems—like proportional representation from a statewide vote—might produce. Jonathan Rodden, in particular, has ably demonstrated that first-past-the-post systems work to the detriment of urban areas due to the high concentration of left-leaning voters in densely populated urban areas.[63] This geospatial sorting leads to left-leaning parties “wasting” proportionally more votes by winning urban districts overwhelmingly. Right-leaning parties, by contrast, waste proportionally fewer votes by winning mostly exurban and rural seats by smaller, but still comfortable, majorities. The result is quite different from what a statewide, proportional allocation of seats would look like.

At the time of Reynolds, this phenomenon had not yet fully developed for a few reasons. First, the two major political parties were far less ideologically coherent and geographically polarized than they are today.[64] Candidates for state legislature or Congress could disassociate themselves from their parties’ national platforms and presidential candidates, while still succeeding electorally.[65] Hence, in the South it was common to find conservative Democrats representing rural areas, while at the same time the Democratic party predominated in northern liberal cities.[66] Similarly, there was a noticeable, even if numerically small, contingent of liberal or “Rockefeller Republicans” representing urban and suburban areas.[67] Before he was mayor of New York City, John Lindsay, for instance, was a Republican Congressman from the Upper East Side’s “Silk Stocking District” and had one of the more liberal voting records in Congress.[68] Even within states, ideological diversity was more common within parties; hence, one-party control of a state, such as that enjoyed by Democrats in the “Solid South,” did not necessarily mean ideological uniformity.[69]

Because of the two major parties’ internal ideological heterogeneity, Reynolds’s mandate of equal representation did not have a clear partisan or ideological impact. By the early 2000s, by contrast, the two most prominent political parties had become far more ideologically coherent.[70] In most states, Democratic candidates are now more uniformly “liberal” or “progressive,” a label that implies certain views on a number of issues such as abortion, firearm safety, environmental protection, consumer protection, and supporting the social safety net more than the opposition party.[71] Similarly, in most states, Republican candidates are now more uniformly “conservative,” a label that implies certain views like support for market-based policies (in some realms), skepticism toward environmental and firearm regulation, relative coolness toward social safety net programs, and, more recently, opposition to immigration.[72]

At the same time, the political preferences of the American public now line up more with geography than ever before. Using presidential preferences as the metric, the 2016 election revealed an incredibly divided nation, with most urban areas strongly preferring the Democratic candidate, Hillary Clinton, most rural areas strongly preferring the Republican candidate, Donald Trump, and suburban areas in the middle.[73]

An electorate that is divided geographically makes no difference on its own in a statewide vote. It is the process of districting that translates these preferences into political power in a particular way. Single-seat, first-past-the-post, compact, contiguous districting almost always leads to a legislature that skews toward the preferences of rural and exurban voters, and this is without any intentional gerrymandering.[74] When intentional gerrymandering is added to the mix, the process cuts even more against the interests of urban residents if the intentional gerrymandering promotes the party that represents the views most associated with exurban and rural voters.[75]

There is strong evidence that this dynamic occurred after 2010 in several key states. The plaintiffs in Gill v. Whitford, a lawsuit alleging intentional gerrymandering of the Wisconsin state legislature, developed a detailed record of partisan gerrymandering after the 2010 wave elections.[76] In Gill, the plaintiffs challenged Wisconsin’s 2011 law, passed by a Republican legislature and signed by a Republican governor, that they alleged intentionally gerrymandered districts so as to ensure a Republican legislative majority for at least a decade, regardless of the statewide cumulative vote.[77] The Supreme Court in Gill ducked the merits of the plaintiffs’ First and Fourteenth Amendment claims, remanding to the trial court for further inquiry into standing.[78] Just one year later, in Rucho v. Common Cause, the Court snuffed out any hope of successfully challenging political gerrymandering under the First or Fourteenth Amendments, ruling five to four that such claims are always nonjusticiable.[79] Nonetheless, while the legal theory may not have been successful, the record and methodology established by the Gill plaintiffs remains very useful for assessing where gerrymandering may have occurred and to what degree.

By way of background, the Gill plaintiffs proposed a three-part test for determining whether intentional gerrymandering had occurred: (1) that a district plan was enacted with partisan intent; (2) that the effect of the plan was to create a large and durable partisan advantage; and (3) that if both of the first two factors are met, defendants must show that the partisan tilt was unavoidable due to the state’s political geography and legitimate districting objectives.[80]

In order to demonstrate the second factor, the Gill plaintiffs relied on a metric known as the “efficiency gap” (“EG”), which is a way of measuring the proportion of “wasted votes” each party experiences in a district-based election.[81] A higher ratio of wasted votes demonstrates that the party has suffered more severely from the legislative districting map. Plaintiffs’ experts extensively analyzed lower-level state legislative elections going back thirty years.[82] This Article will focus on the results from after the 2000 and 2010 redistricting processes, an era of increasing polarization of party politics.[83] While the Gill plaintiffs’ analysis extended only to 2014, the Associated Press (“AP”) analyzed the 2016 elections using the Gill plaintiffs’ same method, and those results are helpful as well.[84] I am not aware of a systematic analysis of EGs from the 2018 state legislative elections available at the time of this Article’s editing, but this Article will incorporate available anecdotal evidence from the 2018 elections. The evidence suggests that many states—especially populous states containing large cities—were gerrymandered in a pro-Republican direction after 2000 and 2010. Later, this Part will analyze five states that exemplify this phenomenon.

A state legislature that veers away from the views of the statewide voters imposes a variety of harms on the population. Such a legislature is likely to pursue policies that do not represent what the voters want. This can be an affirmative harm, in the sense of passing legislation that voters do not support, as well as negative harm: not passing legislation that the voters want.

One might suspect that the governor’s role in the lawmaking process would mitigate the harm that a gerrymandered legislature can impose. Almost every state, after all, elects its governor in a straight-up statewide vote.[85] Particularly where the governor wins a majority or a clear plurality, he or she might be said to have majoritarian bona fides. Governors serve as crucial “veto gates” in the legislative process.[86] Hence, a governor can serve as a majoritarian check on a legislature that strays from popular views.[87]

While there is reason to expect that a governor may serve as a majoritarian check, there is also reason for skepticism. First, with respect to the negative harm imposed by a nonrepresentative legislature, the governor can do little to correct because he or she cannot pass laws on his or her own. A governor may influence the state’s administrative agencies and push them to adopt his or her affirmative agenda, but this is not the same as legislating, and the extent of a governor’s control over the administrative apparatus varies widely by state.[88]

A governor usually has some capacity to block a legislature’s affirmative agenda, but even here the governor’s power will not be absolute. As a repeat player with the legislature, the governor will have incentives to compromise and occasionally allow the legislative majority’s priorities to go through. Even when a governor wields his or her veto pen, the legislature may be so gerrymandered that a veto override is more possible than it would be with more neutrally drawn legislative districts. Further, unlike the federal government, which requires a two-thirds majority to override a presidential veto, several states allow the legislature to override gubernatorial vetoes with bare majorities or with only a three-fifths majority.[89] In addition, while sometimes representative of a majority, governors do not necessarily represent the “median voter,” but rather the triumphant support coalition from the particular election in which they won.[90] Finally, as the recent events in North Carolina and Wisconsin vividly demonstrate, a legislature—even if gerrymandered—can “kneecap” a newly elected governor by passing legislation before he or she takes office to limit his or her powers going forward.[91]

A. Effect of Gerrymandering on Local Governments

If an unrepresentative state legislature only affected state policy, that would be problematic enough. But, as noted above, states retain wide authority to restructure the powers that their local governments possess.[92] Most states have used this power sometime in the twentieth century to create, either by constitution or statute, a system of “home rule” whereby cities—and often counties—could exercise broad policymaking authority in the first instance, subject only to overrule by the state.[93] Most states have retained these systems in theory, but in practice, a raft of preemption and other restrictions on local authority have greatly eroded home rule.[94] Hence, an unrepresentative state legislature may not only decline to pass Proposal X, it may also prohibit any locality in its midst from passing Proposal X.

Such a limitation would have more legitimacy if a majority of the state’s voters credibly opposed the adoption of Proposal X by any level of government. Because most Americans have a healthy respect for local autonomy, it can be safely assumed that the percentage of voters who oppose the adoption of a particular policy by any level of government is smaller than that which opposes the adoption of a policy by the entire state.[95] In many instances, this percentage will be a distinct minority. Nonetheless, a legislature that does not accurately reflect public sentiment is likely to enact a policy that veers from the views of statewide voters.

Gerrymandering can exacerbate this dynamic by leading to a legislature that over-represents the views of a particular slice of the electorate. In many states in the last decade, this has led to an anti-urbanist majority that frequently preempts the enacted ordinances or potential policy priorities of the state’s largest cities.[96] Under Hunter, these acts of the state are perfectly constitutional.[97] When read in light of Reynolds’s embrace of majoritarianism, however, these acts—insofar as the legislature may not accurately reflect the views of the state’s majority due to gerrymandering—are democratically suspect.

Legislative positions on issues veer from the views of voters statewide for a number of reasons besides districting, of course. The effectiveness of interest-group lobbying, for instance, as well as the impact of campaign contributions from interest groups, can play a prominent role.[98] Needless to say, with respect to the contours of local power, these factors play a major role. Recent evidence demonstrates that well-funded organizations, like the American Legislative Exchange Council (“ALEC”), have succeeded at convincing state legislatures to pass a wide variety of preemption legislation aimed at reducing local regulatory authority.[99] This dynamic has recently been more prevalent in states with Republican majorities, although not always.[100] In states with Democratic majorities, interest groups like public-sector unions often wield significant influence and are able to push through legislation that applies across-the-board to local governments.[101]

A detailed analysis of campaign finance’s influence on the relative success of interest groups at different levels of government is outside the confines of this Article. Gerrymandering, however, can exacerbate the influence of money on the political process by leading to a legislature that is already predisposed, even in the absence of campaign money’s potentially distorting effects, to veer away from the views of the majority.[102] Hence, gerrymandered state legislatures are particularly susceptible to attacking urban power and policy priorities. The following instances of preemption from legislatures with strong indicia of gerrymandering are illustrative.

B. State Vignettes

The states highlighted here are those with high efficiency gaps (at least more than 10% in absolute value) in their state houses since 2012 or 2014, as calculated by the plaintiffs’ lead expert in Gill v. Whitford, Dr. Simon Jackman. As articulated by the trial court in Gill, the EG is an objective measurement of the extent to which each of the two major political parties “waste” votes in legislative elections.[103] The EG thus “measures the magnitude of a [districting] plan’s deviation from the [normal] relationship . . . between [statewide] votes and [total] seats.”[104] The Gill plaintiffs proposed 7% as a legally significant threshold for an EG, in part because based on their expert’s testimony, an EG of 7% or more indicates that the districting plan will have tremendous staying power during the decennial period.[105] As made clear by the trial court, a high EG alone does not prove intentional partisan gerrymandering; nonetheless, it may serve as corroborative evidence of an aggressive and effective partisan gerrymander.[106]

In addition to EGs above 10%, this Part focuses on states that also demonstrated other indicia of partisan gerrymandering post-2010, such as a legislative-approved districting map enacted on a largely partisan basis or court decisions finding some partisan gerrymandering in the state districting process. In addition to Wisconsin, these states are (in alphabetical order) Florida, Michigan, North Carolina, and Ohio. Each demonstrates a recent record of aggressively attacking or preempting local authority.[107]

1. Florida

Presidential elections in Florida are famously won by razor-thin margins.[108] Since 2000, a Democratic presidential candidate has won the state twice, while a Republican candidate has won three times.[109] The average margin of victory in the five races is 1.47%.[110] In other statewide races, Florida shows evidence of being a state evenly split between Democrats and Republicans. The last three gubernatorial elections have been extremely close, with Republicans winning by 0.4% in 2018, 1.07% in 2014, and 1.15% in 2010.[111] The 2018 U.S. Senate election was even tighter than any of the last three gubernatorial races, with former Governor Rick Scott defeating incumbent Bill Nelson by 50.05 to 49.93%, or by about 10,000 votes out of more than 8,000,000 cast.[112]

Despite this apparent parity in statewide elections, the Republican party has had a majority in the Florida state legislature since the late 1990s.[113] In recent years, the majority has been utterly dominant, with Republicans enjoying, for instance, an 82-37 seat advantage in the state house and a 26-14 margin in the state senate after the November 2014 elections.[114] There is evidence that Republicans used their control of all three branches of government after both the 2000 and 2010 censuses to ensure that the state legislature (as well as the state’s U.S. House delegation) would become and remain disproportionately Republican. Reviewing the post-2000 state legislative and Congressional districting plans, a federal court observed that “[t]he Republican-controlled legislature intended to maximize the number of Republican . . . legislative seats through the redistricting process, and used its majority power to” effectuate this intent.[115] The Jackman Report showed Florida’s state house to have some of the highest pro-Republican EGs in the country in the 2012 and 2014 elections, with each greater than 10%.[116] The AP report from 2016 shows that in those elections, the Florida state house seat share was expected to be 56.8% Republican under a neutral districting plan, but was in fact 65.8%, for a still-high EG of 9%, one of the highest in the nation that year.[117]

Hence, there is good reason to believe that despite Florida’s politically split electorate, intentional political gerrymandering has played a big role in locking in a Republican majority in the state legislature. This majority, which disproportionately represents exurban and rural areas of the state, aggressively contravened the policy preferences of many—perhaps even a majority—of the state’s voters.[118] For instance, despite President Barack Obama winning the state in 2012 in a campaign largely focused on the merits of his signature domestic achievement, the Affordable Care Act (“ACA”),[119] the Florida legislature rejected attempts to expand Medicaid under the ACA.[120]

With respect to local government in particular, the state legislature has in recent years targeted important subject matters for sweeping preemption. In 2013, the state legislature preempted any city or county in the state from regulating private employers’ workplace benefits and strengthened the state’s ban, initially enacted in 2003, on local minimum wage ordinances.[121] Florida’s intermediate appellate court held that this preemption law prevented Miami Beach from raising its local minimum wage.[122] As further evidence of the legislature’s skew from voters’ views, the voters in 2004 passed a statewide initiative that might be read to preserve the authority of local governments to enact higher minimum wages,[123] an argument rejected by the court that invalidated Miami Beach’s ordinance.[124]

Similarly, in 2016 the legislature preempted localities from regulating polystyrene containers and plastic bags.[125] Some cities have been able to enforce their bans thus far because either the legislature grandfathered them in or because the Florida courts deemed them protected by the state constitution.[126] For most cities in the state, however, the ban will stymie policy experimentation with respect to protecting the environment. In addition, a 2011 law imposes penalties on any locality or local official that attempts to regulate firearms beyond state law.[127] Litigants have attempted to use this law to punish local officials of cities for not affirmatively removing firearms ordinances enacted decades ago from the current city code, even if they went unenforced.[128] Finally, the state legislature recently passed, and the governor signed, an anti-“sanctuary city” law that requires all local jurisdictions to have their police cooperate with federal immigration enforcement.[129] At least one self-proclaimed sanctuary city, South Miami, has sued to block enforcement.[130]

In 2018, despite the extremely close statewide elections,[131] Republicans maintained their utter dominance in the state legislature, with a 73-47 advantage in the house, representing 61% of the seats.[132] From the high EGs demonstrated in earlier years in the decade, these lopsided majorities in the state house likely result to some extent from gerrymandering.[133] The state senate, by contrast, has seen increased parity since a state court ordered the districts redrawn in advance of the 2016 elections.[134] As of September 2019, there are twenty-three Republican senators (as opposed to seventeen Democrats) for 57.5% of the seats.[135] Unlike in many other states, Florida’s forty senate and 120 house districts are drawn independently of each other—i.e., it is not the case that each senate district contains three whole house districts within it.[136]

In sum, the residents of a “purple” state like Florida that might prefer to turn to their local government to enact certain policies have found themselves blocked from doing so due to a political majority in the state legislature that is likely attributable in part to intentional political gerrymandering.

2. Michigan

As demonstrated by the November 2016 election, Michigan is a “swing state.” Donald Trump won the state by a mere 0.22%.[137] The state was reliably Democratic in the six prior presidential elections, although often targeted by Republicans for contention.[138] The state has had two Democratic senators since 2006, while Democrats and Republicans have traded occupancy of the governor’s office in the last two decades.[139]

Despite this seeming partisan parity at the statewide level, Republicans have utterly dominated the Michigan state legislature since the 2010 census, despite losing the statewide popular vote for candidates multiple times. In 2012, for instance, Democratic state house candidates won by 53% to 46% statewide, yet remarkably Republicans held on to a 59-51 majority.[140] Similarly, in 2014, Democratic state house candidates won statewide by a margin of 51% to 49%, yet lost three seats for a 63-47 Republican advantage.[141] Indeed, the Jackman Report estimates the EG for the Michigan state house to be greater than 10% in both 2012 and 2014, among the highest in the nation.[142] In 2016, Democrats and Republicans essentially tied statewide in the house, yet Republicans maintained their sixteen-seat edge.[143] According to the AP, the EG in the Michigan state house in 2016 was the second highest in the nation, resulting in more than eleven excess seats for House Republicans.[144] Most recently, in 2018, Democratic candidates won the state house vote total by 52% to 48%, but Republicans maintained a 58-52 seat edge.[145] Similarly in the state senate, Democrats, in 2018, won 51% of the vote as compared to Republicans’ 49%, yet lost the seat total by a margin of 22-16.[146]

As demonstrated by the Jackman Report, high pro-Republican EGs date back to the 2000s.[147] This is not surprising, given that after both the 2000 and 2010 censuses, the Republican-controlled Michigan legislature adopted partisan districting plans with the approval of Republican governors.[148]

The skewing of the Michigan legislature away from statewide voter preferences has resulted in legislation that deprives local governments of significant authority. Most notably, in 2012, with Detroit’s fiscal crisis in the background, the state legislature enacted an emergency manager law that stripped elected city councils and mayors of their powers despite a statewide initiative passed a month earlier that sought to rescind that law.[149] In other words, despite the Michigan voters’ clear statewide preference for protecting local democracy, the legislature reached a very different conclusion. In addition to engendering bitterness among some Detroit residents over their city being steered into bankruptcy by a state-appointed functionary, the reinstituted emergency manager law may also have played a role in causing the Flint lead-poisoning water crisis as Flint too was placed under state-imposed emergency management.[150]

In the last few years, the Michigan legislature has also enacted aggressive deregulatory preemption. In 2015, the legislature enacted the so-called “Death Star” bill preventing cities and counties from regulating any aspect of the employment relationship, including minimum wage, leave, and benefits.[151] In 2016, the state legislature enacted a ban on plastic bag bans,[152] after the commissioners of Washtenaw County, which includes the city of Ann Arbor, voted in favor of an ordinance that would impose a ten-cent fee on paper and plastic bags.[153] The new state law prevented the proposed local ordinance from taking effect.[154]

Michigan, therefore, is a clear example of a state whose legislature does not accurately reflect the views of its voters statewide, and intentional partisan gerrymandering appears to be a major culprit.

3. North Carolina

Perhaps in no state were the effects of gerrymandering as devastating to local governments—and, at least for a short while, the economy of the entire state—than in North Carolina. North Carolina was traditionally a politically moderate state. While it often voted Republican in presidential elections, it had a tradition of “moderate-to-progressive state government” and electing senators and governors from both major political parties.[155] In 2008, North Carolina became a true “swing state” in presidential elections by voting for the Democratic candidate—Barack Obama—for the first time since 1976.[156] In the two subsequent elections, 2012 and 2016, the state voted for the Republican presidential candidate by an average margin of 2.85%.[157]

After a Republican sweep of both houses of the legislature and the governor’s office in 2010, however, the state legislative maps were drawn in an entirely partisan fashion. Indeed, the vote in favor of the 2011 redistricting plan in the state house was 66-53, with all but two Republicans voting yes and all Democrats present (one was absent) voting no.[158] As a result of this apparent intentional gerrymander, the North Carolina legislature took on a profile that skewed sharply away from the state’s traditional approach to governance. As the Jackman Report demonstrates, the North Carolina state house’s EGs in 2012 and 2014 were among the highest in the nation in absolute value, with each greater than 10%.[159] The AP’s analysis of the 2016 state house elections found a slightly lower, but still significant, EG of 5.51%, resulting in an estimated additional 6.6 Republican seats.[160]

In 2016, a federal district court found in Covington v. North Carolina that the state legislature’s 2011 districting plan violated the Equal Protection Clause by engaging in undue racial gerrymandering.[161] After a tortuous litigation history, the Supreme Court affirmed the redrawing of several state house and senate districts before the 2018 elections.[162] With these new district maps in effect for the November 2018 elections, Democrats reduced the Republican majority in the state senate from 34-15 to 29-21,[163] and in the house from 75-45 to 65-55.[164] Democrats’ gains in 2018 were significant because they denied Republicans the three-fifths supermajority necessary to override gubernatorial vetoes.[165]

While there is not yet an EG analysis available regarding the 2018 North Carolina state legislative elections, preliminary tallies show a continued advantage for Republicans at converting votes into seats, despite the redrawing of some districts after Covington. Democrats won slight majorities of the statewide vote for candidates of both chambers, yet ended up with a minority in each.[166] Covington, of course, addressed only racial gerrymandering and not political gerrymandering.

The apparent gerrymandering of the North Carolina General Assembly has caused the legislature’s majority to skew away from the preferences of the state’s urban voters. Hence, the legislature has been aggressive since 2012 in preempting the priorities of urban centers within the state. North Carolina’s cities derive their powers from statute rather than from the state constitution and are thus particularly susceptible to being overridden by the state legislature.[167]

In one of the most nationally prominent instances of preemption in recent years, the legislature in 2016 preempted Charlotte’s ordinance that sought to extend the antidiscrimination protections of local law to gay, lesbian, and transgender persons.[168] The state law, popularly known as House Bill 2 (“HB2”) or “the bathroom bill,” because its supporters argued that it would keep men out of women’s bathrooms,[169] spurred a national backlash leading numerous companies, sports organizations, and high-profile entertainers to boycott the state.[170] In addition to preempting local antidiscrimination laws, HB2 preempted the ability of local governments to enact higher minimum wage ordinances, regulate leave or benefits, or require that city contractors hire local employees.[171] The national outcry over HB2 ultimately forced the legislature to pass a partial repeal.[172] The repeal, however, prohibited any political subdivision from regulating private employment practices or public accommodations until 2020.[173]

Moreover, to ensure that they maintained their grip on power, the Republican legislative majority stripped powers from the governor’s office after the Democratic candidate, Roy Cooper, won the November 2016 election.[174] In 2018, the Republican majority in the legislature referred constitutional changes to the voters that would have further reduced the governor’s powers,[175] but voters rejected these proposals statewide.[176]

4. Ohio

Ohio is a politically competitive state, commonly understood as the nation’s most reliable bellwether in presidential elections.[177] While 2016 had a relative wide margin for the winning candidate, Donald Trump—51.69% to 43.56%—the average margin of victory over the course of the four prior presidential races is 3.29%.[178] Since 1992, moreover, Democratic presidential candidates have won the state four times, while Republican candidates have won the state three times.[179] Other statewide races likewise reflect Ohio’s moderation. Republican John Kasich won the governorship by only 2% in 2010,[180] his Republican successor Mike DeWine won by less than 4% in 2018,[181] and the state’s two U.S. Senate seats are split between a Democrat and Republican.[182]

Despite this clear evidence of balanced partisan competitiveness in statewide elections, the Republican Party has dominated the Ohio General Assembly for decades. The state senate has been controlled by Republicans in an unbroken streak since 1992, with increasing margins that are currently peaking at twenty-four Republicans to nine Democrats.[183] In the state house, since 1994 the Democratic Party has only had a majority in one session—2008 to 2010—and the current partisan divide is a 61-38 Republican majority, down from a 66-32 majority from before the 2018 elections.[184]

The Jackman Report showed that Ohio’s state house was among the top ten states in pro-Republican EG in both the 2012 and 2014 elections.[185] Similarly, the AP’s study of the 2016 state house elections ranked Ohio’s EG among the top ten nationally, and the eighth highest among those that favor Republicans, resulting in an excess of over five state house seats for Republicans.[186] In 2018, Republicans won 52% of the cumulative vote for the state house, but 63% of the seats; they won just 48% of the statewide vote for the seventeen senate contests on the ballot, but won 65% of the seats.[187]

The evidence strongly suggests that intentional political gerrymandering has played a significant role in supporting the continuing Republican majority in the Ohio General Assembly. This majority, which as in many similar states has disproportionately represented exurban and rural areas in the state,[188] has aggressively contravened the policy preferences of the state’s voters living in the state’s larger cities.

With respect to local government in particular, the state legislature has targeted important subject matters for sweeping preemption. In 2002, the General Assembly preempted home-rule authority for cities to respond to serious local problems involving predatory lending.[189] In 2006, the General Assembly preempted local authority over residency for city employees[190] and removed longstanding home-rule authority to regulate gun safety.[191] The General Assembly has likewise preempted the ability of cities to enact local-hire laws.[192] The General Assembly has also limited the ability of cities to use red-light and speed cameras.[193]

In sum, residents of a quintessentially “purple” state like Ohio who prefer their local governments to enact policies that match their “small-d” democratic preferences find themselves repeatedly blocked by an entrenched political alignment in the state legislature that may well have been attributable to intentional political gerrymandering.

5. Wisconsin

The district court in Whitford v. Gill examined Wisconsin’s record of gerrymandering after 2010 in detail.[194] After concluding that the legislature’s 2011 redistricting law, Act 43, was infused with clear partisan intent,[195] the court analyzed the effectiveness of the effort to ensure Republican dominance of the legislature through districting:

In 2012, the Democrats garnered 51.4% of the [cumulative state assembly] vote, but secured only 39 seats in the Assembly—or 39.3% of the seats. In 2014, the Democrats garnered 48% of the vote and won only 36 seats—or 36.4% of the seats.[196]

The district court in Gill noted that as calculated by a plaintiff’s expert, the pro-Republican EGs for 2012 and 2014 were 13% and 10%, respectively.[197] Stated differently,

[T]he Republican Party in 2012 won about 13 Assembly seats in excess of what a party would be expected to win with 49% of the statewide vote [under a more neutral districting plan], and in 2014 it won about 10 more Assembly seats than would be expected with 52% of the vote.[198]

Later elections provide further evidence of Act 43’s effectiveness at ensuring a Republican seat majority regardless of popular sentiment. The EG for the 2016 assembly races was just under 10%, the third highest in the nation that year, which likely translated into nine or ten excess Republican seats in the Wisconsin House.[199] The EG for 2018 was a whopping 15%; as a result, despite Democrats receiving 54% of the total votes cast for Assembly candidates in 2018, Republicans maintained a 63-99 majority.[200] This occurred in a year in which Democratic candidates won all four statewide offices on the ballot.[201] A minority, therefore, has successfully entrenched itself in the legislature in Wisconsin. Because this legislative minority draws disproportionate strength from smaller towns and rural areas, it has used its power to target policies preferred by the residents of the state’s large, Democratic-leaning cities like Milwaukee and Madison.[202]

Since 2012, the state legislature has attempted to erode local policymaking authority through aggressive preemption. A list of subject matters preempted by the state legislature since 2012 include:

* nutrition and food policy;[203]

* issuance of photo identification cards by local governments, particularly to prohibit their use in voting;[204] and

* municipal employee residency requirements.[205]

To be sure, other key preemptive laws in Wisconsin were passed before the apparent gerrymandering took effect, such as a state law in 2005 preempting local minimum wages,[206] and a 2011 law preempting local paid sick leave ordinances.[207] The inability to reverse these laws, however, may also be partly attributable to gerrymandering.[208]

C. Wrap-up

The five states discussed above are not the only ones whose state legislatures have demonstrated indicia of intentional political gerrymandering in recent years. They are simply those with the clearest and most egregious records in that regard. Other states with notable pro-Republican gerrymandering within the last decade include Indiana, Virginia, and New York.[209] In Indiana, Republicans controlled the legislative districting process completely after 2010.[210] In Virginia, the districting processes were more bipartisan, but still led to pro-Republican results until 2019,[211] when Democrats flipped both houses of the state legislature under redrawn districts that neutralized the prior Republican advantage.[212] In New York, there was a tradition of each house deferring to the others’ drawing of its own districts, thus allowing the Republicans to gerrymander the senate and control it until 2018.[213] States with indicia of pro-Democratic gerrymandering after 2010 include Colorado, Maryland, Nevada, and Rhode Island.[214] Of these four, Rhode Island’s record is the most glaring.[215]

If gerrymandering leads to preemption, as this Part has suggested, then perhaps the solution to such preemption is to focus on the districting process rather than on strengthening home-rule protections for local governments, as some might suggest. This is a reasonable conclusion to draw from the discussion, and this Article’s goal is not to argue for a particular doctrinal solution to the problem of democratically suspect preemption.

Nonetheless, for those who are concerned with democratically suspect preemption, there are reasons why districting reform itself may not solve the problem, or may lack strength as a tool. First, while districting can create more democratically legitimate state legislatures going forward, it is difficult for future legislatures to undo all the preemption of years past. Legislation can be “sticky,” as groups with a vested interest in preemption can be expected to mobilize to block a repeal.[216] Second, with respect to litigation, while most states have some sort of home-rule provision that might protect local democracy, the textual “hook” for challenging gerrymandering in state constitutions is often less clear.[217]  In a few states, however—namely, Florida, North Carolina, and Pennsylvania—litigants have successfully challenged political gerrymandering under the states’ constitutions.[218]

Regardless of what the best solution is, the problem is pellucid: gerrymandered legislatures lead to legislation that does not accurately represent public views. Moreover, in recent decades, when intentional partisan gerrymandering favors Republicans, it exacerbates the pre-existing anti-urban bias inherent in first-past-the-post, single-member districts. While the disconnect between voter views and the views of their legislative representatives that results from intentional and unintentional gerrymandering is a significant, first-order democratic harm, the second-order harm imposed by preemption is also significant. In other words, if the majoritarian legitimacy that the Supreme Court sought to imbue in legislatures through Reynolds is lacking, the state’s claim to the awesome Hunter power—particularly when used to deprive local governments of pre-existing powers—is more democratically suspect.

III. Direct Democracy as Unfiltered Voice of the “Majority?”

A potential antidote to an unrepresentative legislature is the initiative process, or direct democracy. In states that have it—approximately twenty[219]—voters can by plebiscite correct state legislation that deprives local governments of power. As noted previously, in Michigan, voters passed a referendum to restore local control of city governments after the state instituted an aggressive emergency manager program after the Great Recession.[220] The legislature, however, reversed this decision by legislation in the next session.[221]

The possibility of direct democracy serving as an unfiltered voice of the people has been appealing since the process emerged in the late 1800s.[222] The stated hope of its early Progressive supporters was that it would allow “the people” to circumvent a legislature beholden to special interest groups.[223] On issues like medical or recreational marijuana, a higher minimum wage, and stricter gun control laws, proponents have frequently bypassed legislative gridlock to change the law by direct vote.[224]

On the other hand, one must not be too glib in describing the benefits of direct democracy. There is an entire literature devoted to its shortcomings.[225] Oregon Supreme Court Justice Hans Linde famously argued that the initiative process, when it targeted distinct minorities, such as gays and lesbians, might violate the “Guarantee Clause” of the federal Constitution.[226] In the racial context, scholars and advocates have argued that the initiative is a blunt tool of majority oppression.[227] Sensitive to these concerns, the Supreme Court in the early 1980s articulated in the direct democracy context a “political process doctrine” of the Fourteenth Amendment’s Equal Protection Clause, whereby state decisions that systematically weaken the power of racial minorities may be invalidated.[228] Moreover, in the fiscal realm, critics have blamed direct democracy for hamstringing the functioning of state and local governments after voters in several states adopted constitutional supermajority requirements for, and other restrictions on, revenue-raising.[229]

Nonetheless, direct democracy at least appears to be the voice of the people and, unlike a gerrymandered state legislature, it strictly comports with the principle of one-person, one-vote. It is useful, therefore, to assess preemption and other attacks on local authority that may result from the initiative process through the lens of some of the primary critiques of direct democracy. In addition to those mentioned above, other critiques of direct democracy include:

*  It does not lead to outcomes more closely aligned with public sentiment.

*  It leaves questions to citizens who are not able to make good decisions.

*  It amplifies the power of interest groups.

*  It is, ironically, less democratic than representative government.

This Part will elaborate on these major critiques of direct democracy. It will then analyze these critiques as they apply to some specific instances of direct democratic preemption, or resistance to preemption, of recent note. These include minimum wage, voting to protect local democracy in Michigan, and the sanctuary state policy in Oregon. Without offering any firm conclusions, this Part suggests that direct democracy complicates the preemption analysis. Depending on which critiques of direct democracy are of most concern, preemption via plebiscite may be more or less troublesome than its legislative counterpart.

A. Oppression of Minorities

Because it relies on a raw majority vote, direct democracy has long been criticized for being a means for the majority to run roughshod over minorities, whether racial, ethnic, sexual, or other. Indeed, various initiatives have been seen as examples of this type, leading to high-profile litigation before the United States Supreme Court. Prominent examples include Washington’s Initiative 350, passed in 1978 to invalidate Seattle’s school busing plan, and overturned by the Court in Washington v. Seattle School District No. 1, in 1982,[230] and Colorado’s Amendment 2, which invalidated three cities’ ordinances that established sexual orientation as a basis for illegal discrimination.[231] The Court rejected Amendment 2 in Romer v. Evans.[232]

Most of the academic commentary has classified Seattle School District No. 1 and Romer as civil rights cases, but as Judge David Barron noted, they are also cases about protecting local democracy.[233] Interestingly, Seattle School District No. 1 and Romer are not about protecting local “minority-majority” populations.[234] Rather, they are about protecting localities with visions regarding integration and the treatment of minorities that differed from those of voters statewide.

In Seattle School District No. 1, a school board representing a majority-white community[235] adopted a “voluntary” school integration plan.[236] Through Initiative 350, Washington’s voters, also majority white, preempted this local prerogative.[237] The Supreme Court invalidated the initiative under the Equal Protection Clause of the Fourteenth Amendment. In striking down the initiative, the Court noted that the initiative “remove[d] the authority to address a racial problem—and only a racial problem—from the existing decisionmaking body, in such a way as to burden minority interests.”[238] Had the Seattle School Board repealed the busing policy itself, that may have been perfectly constitutional.[239] The constitutional problem with Initiative 350, according to the Court, was that it “lodg[ed] decisionmaking authority over the question at a new and remote level of government.”[240]

In Romer, three Colorado cities—Aspen, Boulder, and Denver—adopted antidiscrimination ordinances that included sexual orientation as a protected class.[241] By constitutional initiative, the state’s voters overruled these cities’ ordinances.[242] While Aspen, Boulder, and Denver may have had higher gay populations than the rest of Colorado, they were nonetheless presumably majority-heterosexual cities, just like the state as a whole. As is often the dynamic, voters in politically liberal big cities or college or resort towns took a more progressive view toward homosexuality than voters statewide.[243] The Supreme Court invalidated Colorado’s Amendment 2, again under the Equal Protection Clause. The Court applied something akin to rational basis review, holding that Amendment 2 was motivated by “a bare . . . desire to harm a politically unpopular group” and that this desire could not “constitute a legitimate governmental interest.”[244]

Although formally an Equal Protection case, Judge Barron argues that Romer also stands for the proposition that “localism” can advance “positive constitutionalism” when “diverse communities . . . unite to promote constitutional values that their states have ignored.”[245] Romer may also demonstrate the Court’s discomfort with the initiative process as the means of displacing a locality’s ability to promote equality. In this sense, Romer brings to mind Justice Linde’s argument as to why the initiative process is potentially antirepublican.[246]

The Court’s rationale in Seattle School District No. 1 has become known as the “political process doctrine”—i.e., state actions that restructure the political process to the detriment of minority voters are suspect under the Fourteenth Amendment’s Equal Protection Clause. Although the Court would later cast doubt on this doctrine in Schuette v. BAMN, it remains a theoretically viable doctrine.[247] It should be noted that the Court has only invoked the doctrine in the racial context; Romer, although concerned with (sexual-orientation) minorities, did not explicitly rely on the “political process” doctrine. Hence, preemption by initiative may trigger federal constitutional concerns, particularly when it strips local governments of power to act in areas where they may promote a vision of “positive constitutionalism,” as Barron notes.[248]

Other instances of preemption arguably bolster minority rights. For instance, in several “sanctuary” states, state law prohibits state and local government officials from cooperating with federal authorities to remove persons lacking lawful immigration status.[249] Oregon is such a state, with the legislature having passed a sanctuary statute in 1987.[250] In 2018, opponents of the policy put an initiative on the ballot to repeal the state law.[251] Civil rights, Hispanic advocacy, and most progressive groups ardently opposed the initiative.[252] The initiative would have restored local control over immigration, but in a way that could be seen as limiting civil rights as well as curbing “positive constitutionalism,” broadly defined.[253]

Insofar as one might view the sanctuary state repeal as anti-Latinx, Oregon’s experience demonstrates that preemption by initiative need not always harm minorities. Indeed, an overwhelmingly white state, by a vote of 63% to 37%,[254] supported a policy that likely had no direct effect on the vast majority of residents, who are either citizens or white, and therefore unlikely to experience the racial profiling that some thought might result from a repeal.[255] A breakdown of support for the measure demonstrates that it was largely the urban, more densely populated, and, in some cases, more ethnically diverse counties that opposed a repeal.[256] The more rural counties, some of which were less Hispanic than the rest of the state, but some of which were more so, generally supported the repeal.[257]

From these examples, therefore, one can observe that preemption by initiative has a mixed track record from the perspective of minority oppression. It has been used to limit local expansion of civil rights, but recently it has also been used to retain statewide protections for vulnerable groups, as in Oregon.

B. Unrepresentative of Desired Public Opinion

A second critique of the initiative is at first blush ironic: it is less democratic than representative lawmaking. While a direct up-or-down vote might seem like the best possible gauge of the public’s views, some commentators have argued that the situation is not as simple as it seems. This critique can be roughly divided into two subcamps. First, some scholars—notably, Sherman Clark—have argued that the crude, binary nature of the initiative process deprives voters of the ability to express their priorities among issues.[258] People feel as if they have significant input, but in fact the inability to express preferences among issues limits people’s effective use of political power.[259] Relatedly, Ethan Lieb has documented how direct democracy’s reliance on “naked” preferences “is potentially troublesome because it makes little effort to educate citizens on the issues upon which they are voting and gives them no well-suited forum to deliberate about those issues.”[260]

To be sure, in states that have the initiative process, votes thereunder do not take place in a vacuum. Civic organizations and others often host forums for advocates and opponents to debate proposed initiatives. Newspapers publish columns and letters to the editor on the subjects. Moreover, some states mandate more deliberation before initiatives go on the ballot, such as by requiring a citizen commission to review proposed ballot measures.[261] But for Clark and Lieb, these opportunities pale in comparison to the normal give-and-take over a legislative proposal.[262] Hence, even when voters vote for something they think they support, they may end up dissatisfied with the result because their preferences were formed through a defective process.

A related critique of direct democracy focuses on the demographics of voters who participate in direct democracy. David Magleby has demonstrated that lower-income and less-educated voters, while already less likely to vote overall, are more likely to skip voting on ballot questions.[263] Even on issues of direct relevance, Magleby notes that low-income voters report leaving their ballot blank because the measure in question is “too long and hard to understand.”[264] Hence, direct democracy may exacerbate an already-existing bias in our political system toward representing the views of wealthier and more highly educated voters.[265]

The net result of these critiques is that while the outcome of a ballot initiative may reflect voter preferences in some barebones form, this is not the sort of preference our political system should privilege. In other words, because the process is flawed, it is possible that voter initiatives will produce outcomes that are out of step either with naked public opinion (due to demographic skew of voters) or with the more informed, deliberate public opinion preferred on normative grounds.

With respect to preemption, these critiques of direct democracy are most applicable in the context of a successful initiative that imposes preemption, rather than one that rejects preemption. A particularly good example of such an initiative would be California’s famous, or notorious, Proposition 13 (“Prop 13”).[266] Adopted in 1978, this constitutional amendment made it much more difficult for the state as a whole to raise taxes.[267] It also significantly capped the ability of local governments to raise taxes.[268] Prop 13, therefore, amounted to an extremely strong dose of fiscal preemption, severely limiting the ability of cities and counties to decide for themselves what level of taxation and services they would prefer.[269]

Prop 13 and other examples like it might be seen as paradigmatic examples of what the Clark critique illustrates. Everyone wants lower taxes if at no cost, but many do not want to make the tradeoff for lower services. When phrased as an up-or-down ballot initiative, however, voters are not asked to weigh the potential service cuts when voting. Hence, because there is no give-and-take with other governmental priorities, the tax cut wins out and services are cut later as an unforeseen consequence. Indeed, surveys after Prop 13’s passage indicated that voters inaccurately thought they could “benefit from its cut in property taxes without a significant reduction in the level of government services.”[270]

If these unintended consequences happened just at the state level, that would be problematic enough, but measures like Prop 13 also preempt local control. The preemptive aspect of the initiative, however, may not be front-and-center to the voting public. Rather, the public is more likely to perceive that the initiative is about taxes generally, not realizing that local control is also very much on the ballot. Of course, opponents of Prop 13—and other measures like it—were free to make the argument that the initiative would preempt local control. Amidst the cacophony of voices arguing in favor or against the initiative, however, the distinct angle of local control may attract little attention. Indeed, the campaign against Prop 13 focused on a reduction in government services generally without emphasizing the local angle.[271] One technique that might help would be a prominently placed special statement in the voters’ pamphlet statement—or perhaps even on the ballot—describing the initiative’s expected impact on local authority. To varying degrees, some states have used a similar technique with respect to the fiscal impact of ballot measures.[272]

Just as preemption may be underestimated by voters when passing an initiative imposing a new statewide policy, voters may also not fully consider the implications for local authority of repealing a state law. Oregon’s recent experience with Measure 105, as discussed above, is illustrative.[273] The text of the ballot measure and arguments in support were clear that it would repeal a statewide preemptive law.[274] If that had happened, it is likely that the jurisdictions representing a substantial percentage—or perhaps a solid majority—of the state’s population would have continued to prohibit their law enforcement officials from enforcing federal immigration law. Indeed, Multnomah County, with over 800,000 people and almost twenty percent of the state’s population in 2016, passed a sanctuary resolution that endorsed the county sheriff’s continued compliance with Oregon’s sanctuary law.[275] Although not self-executing if the state statute had been repealed, the resolution indicated the county’s commitment to protecting immigrants from federal enforcement. Other populous counties, like Clackamas, stopped short of adopting sanctuary policies knowing that they were unnecessary given the state law on the books; Clackamas, however, passed a different resolution that signaled its openness to protecting immigrants regardless of legal status.[276] The message opposing Measure 105, of course, did not stress the probability that repeal of the state statute might still effectively leave much protection in place for undocumented immigrants. Rather, the successful opposition to measure 105 focused on the either-or proposition of enforcing federal immigration law in Oregon or not.[277]

For those concerned with local control qua local control, the resounding defeat of Measure 105 was a significant loss. Repealing the sanctuary law would have forced elected community leaders at the local level to decide for themselves whether to enforce federal immigration law. To be sure, many of these officials took public stands on the statewide measure, so the process of debating Measure 105 helped reveal their preferences on the larger issue of immigration.[278]

On the other hand, for those who generally value local control yet voted against Measure 105, a clearer explanation of the preemptive effects of retaining the current “sanctuary law” may not have made a difference. All advocates of local control prefer some statewide “floors” below which local authority may not dip. Such floors include basic statewide health and safety standards, minimum wages, and antidiscrimination laws. The opponents of Measure 105, therefore, may have viewed freedom from state and local immigration enforcement as the kind of vital civil right that ought to be provided at a “floor” level throughout the state, regardless of the effect on local choice.[279]

C. Amplification of the Power of Interest Groups and the Wealthy

A third oft-cited critique of direct democracy is that it amplifies the power of interest groups, particularly through their expenditure of money to get initiatives on the ballot and/or to promote them or defeat them.[280] Interest groups, of course, are powerful at all levels of government and in all spheres, from city councils to administrative agencies to Congress. They derive their power in large part from their roles in the interrelated processes of endorsing and funding candidates in elections, motivating voters, and then lobbying candidates once in office to pass their preferred legislation or to oppose their disfavored legislation. With respect to ballot initiatives, those two activities are compressed into a single election that determines the state’s policy position on an issue. The relationships that are so important to lobbying are much less relevant in the context of direct democracy since those in favor or against are speaking directly to the voters rather than attempting to woo a relatively small group of legislators.[281]

Hence, for those promoting an initiative (assuming it was not referred by the legislature), there are the “up-front costs” of getting an initiative on the ballot, which would require more traditional legislative lobbying. These costs potentially include hiring attorneys to draft the proposed language, gathering the requisite number of signatures, and then again hiring attorneys to litigate the final ballot language if the state authorities—department of justice, secretary of state, or the courts—issue a ballot title that proponents find unfavorable, or perhaps even refuse to certify the ballot measure at all.[282] Beyond the up-front costs, there are the potentially huge sums of money to advocate for an initiative’s passage or defeat. The most expensive campaigns include significant television advertising campaigns and cost in the tens of millions.[283] In addition, an effective “ground game” to get out the vote requires significant expenditures.

David Damore and Stephen Nicholson note that the ballot initiatives most likely to mobilize voters are those that “tap into deep divisions in contemporary American politics, such as social issues and tax increases.”[284] More technical or esoteric statutory or constitutional “fixes” referred by the legislature are less likely to excite the general public and inspire a well-funded campaign. For the issues that do arouse voters and attract interest-group spending, it is widely accepted that it is more difficult to buy a “yes” on an initiative from voters than it is to buy a “no.”[285] Elisabeth Gerber has demonstrated that well-funded interest groups, in particular, are adept at blocking ballot initiatives, but less successful in pushing initiatives that they promote.[286] Hence, to the extent that we are concerned with interest group strength, it will be more pronounced when there is an imbalance, with more well-funded interest groups lined up on the “no” side of an initiative. As demonstrated in the table below, with respect to preemption this dynamic will be particularly acute when an initiative seeks to restore local power that was previously preempted, whether by initiative or ordinary legislation.


Ballot Measure Campaign In Favor Against
Preempt More Difficult Easier
Restore Local Control Most Difficult Easiest


The recent effort in California to repeal the statewide preemption of rent control is a case in point. The state legislature put significant restrictions on cities’ authority to enact rent control in 1995 through the Costa-Hawkins Rental Housing Act.[287] The 2018 voter-initiated Proposition 10 (“Prop 10”) sought to repeal the prior preemption.[288] The opposition to Prop 10 received much stronger financial support, particularly from the real estate industry, outraising the proponents by a margin of three-to-one and spending $76 million.[289] A well-funded interest group, therefore, was well-positioned to buy a “no” with respect to Prop 10, which the voters defeated by an almost 60% to 40% margin.[290]

One reason to suspect that money plays a greater role in the initiative process than it does in the legislative process is that there are fewer limitations on contributions than there are with respect to legislative or executive candidates. Hence, while in many states a political action committee or wealthy individual might be limited to a few thousand dollars per candidate, in the same states there may be no limit on contributions to a ballot initiative campaign.[291]

To the extent that this critique is concerned with money’s influence on the plebiscite, it is worth distinguishing among interest groups and wealthy individuals promoting their own views. The phrase “interest groups” describes, at a general level, a “voluntary association[] independent of the political system that attempts to influence the government.”[292] Interest groups include groups from the for-profit sector, like the Chemical Manufacturer’s Association or the Mortgage Bankers Association.[293] They also include nonprofit or governmental entities like the National Association of Counties or the National Association of Independent Colleges and Universities, as well as labor unions.[294] They also include ideologically motivated groups, sometimes referred to as “advocacy groups,” such as NARAL Pro-Choice America or the National Rifle Association, that promote a consistent position on a particular issue or set of issues.[295]

Wealthy individuals unaffiliated with any particular interest group may contribute to campaigns and organizations because of their own personal views on important issues. Their power may be relatively magnified in the initiative context because money is less regulated, and the organizational strength that interest groups have may be comparatively more valuable in processes—such as contribution-limited elections and lobbying—in which manpower and voter communication matter more. Of course, wealthy individuals may also form or fund more traditional advocacy organizations when they find a cause of particular concern.

With respect to preemption by initiative, with the exception of organizations of cities, counties, or special districts, there are not currently any interest groups that care about local power per se. Hence, almost any interest group can be expected to support legislation that preempts local control if it furthers their substantive agenda. It is always possible that a “white knight” billionaire might care particularly about local control, but to date no such savior of local democracy has emerged.

Where an initiative promotes an interest group’s agenda at the expense of local control, therefore, we should be particularly skeptical of the extent to which it accurately represents voters’ views when the proponents are especially well-funded. The recent attempts in Washington and Oregon to ban soda taxes are a case in point. The motivation for the proposals was primarily to blunt the adoption of soda taxes at the local level as a small but increasing number of cities have begun adopting them.[296] The City of Seattle adopted a soda tax in 2017.[297] In Oregon, no jurisdiction had yet adopted a soda tax, but it was being considered in Multnomah County, the state’s most populous county and home to its largest city, Portland.[298]

The Washington measure was a statutory initiative (Washington has no constitutional initiative)[299] and it exempted any pre-existing soda taxes, which had the effect of preserving Seattle’s tax.[300] The measure passed by a margin of 56% to 44%, with proponents of the measure—largely the soda industry—outspending opponents by a factor of 178 to 1![301] With Coca-Cola alone contributing over $10 million, the pro-soda forces raised and spent over $20 million.[302] A handful of public-interest oriented groups barely mustered over $100,000 in opposition.[303] Hence, while it may be more difficult to buy a yes than a no, such a disadvantage may be overcome with lopsided financing.

Oregon’s Measure 103, in contrast to Washington’s, was a constitutional initiative.[304] It failed 57% to 43%.[305] Proponents spent about $8 million while opponents spent about $11 million.[306] Hence, in Oregon, the financial backing for each side was much more balanced than in Washington, and even favored the opposition, which may well have affected the outcome.[307] Coca-Cola and Pepsico did not spend anywhere near the amount of money in Oregon that they spent in Washington. This disparity may have been due to a handful of factors. First, the existence of at least one soda tax in Washington, as opposed to none yet in Oregon, may have heightened the urgency supporters of a Washington ban felt for their cause. Second, Oregon’s Measure 103 supporters may have misfired by seeking a constitutional, as opposed to statutory, change, which made it easier for opponents to paint the measure as a straitjacket.[308] Third, a variety of progressive groups organized around opposing Measure 103, whereas in Washington the “no” coalition was much weaker.[309] Finally, once polling started indicating that Washington’s measure had a better chance of passage, it likely attracted comparatively more money as election day approached.

For those concerned with local democracy, in particular, the anti-Measure 103 message in Oregon seemed to resonate. Local editorials picked up on the theme.[310] The mass-produced signs by progressive groups labeled it as “anti-democracy.”[311] If supported by enough money and organizational clout and connected to other important messages, then, a message partly rooted in protecting local democracy may help defeat initiatives that erode it.[312] Campaigns opposing threats to local control will likely stand a better chance than those measures that seek to restore local control, particularly when the latter, as in the case of California’s Prop 10, are up against a well-funded opposition.

In conclusion, the initiative process is a mixed bag with respect to local democracy. It has frequently subverted local democracy, and in ways that have undermined minority rights, but it has also been used to protect minority rights throughout the state. Moreover, anecdotal examples from Michigan and Oregon demonstrate that the issue of local control can resonate with voters in initiative campaigns, particularly if supported by an energetic campaign and linked to other important substantive political messages.

IV. Other Criteria: Horse Trading and Punitive Preemption

Finally, other criteria that are potentially relevant to judging the normative merits of the process leading to preemption are the presence of “horse trading” as well as whether the final preemptive product contains a punitive element.

A. Horse Trading

With respect to horse trading, the interest groups pushing for policy change at the local level may be willing to sacrifice local autonomy in a particular sphere in exchange for a state standard that is higher than the pre-existing floor but lower than what they were capable of achieving from the most-friendly city council. Consider the trajectory of earned sick leave in Oregon. Proponents were successful in persuading the Portland and Eugene city councils to adopt a paid sick leave policy when the state had none.[313] This success placed the issue on the state legislative agenda, as business groups understandably became concerned about a patchwork of local approaches and were willing to compromise on a statewide uniform measure.[314] Ultimately, the state legislature adopted a statewide standard for paid sick leave that was not as generous as Portland’s and that preempted cities around the state from implementing their own Portland-like measures.[315] Nonetheless, by adopting at least some paid sick leave policy statewide, the measure was widely supported by the progressive groups that initially promoted the policy at the local level.

As the Oregon paid-sick-leave example—and many others like it—show, when preemption results from a legislative bargain, it may represent the voluntary relinquishment of local power by interest groups and their local elected allies in order to advance a policy statewide. As such, democratic legitimacy concerns about preemption in such instances are less acute, particularly when the interest groups relinquishing control and their city allies proceed from a stance of credible bargaining power.[316]

To be sure, there were “losers” in Oregon’s sick-leave tradeoff. Several cities and counties were forced to comply with paid sick leave when they had no inclination to adopt such a policy on their own.[317] Insofar as one might have been concerned with the loss of the ability to innovate in a pro-sick-leave direction, however, the sacrifice of cities’ ability to further innovate in that regard appears to have been a well-informed bargain by proponents of the policy. Advocates must be careful about sacrificing the ability to innovate, however, given that the demands for policy change can be unpredictable.

B. Punitive Preemption

What to do about punitive preemption? The landscape of preemption has changed dramatically in the last decade.[318] The concern is no longer simply depriving a local government of a power previously exercised by it, but rather the threat of imposing penalties upon local governments and officials for pursuing policies deemed antithetical to state law. Should we be less concerned about punitive preemption if it results from a majoritarian legislature or voter-enacted initiative as opposed to a gerrymandered legislature, for instance? I would argue yes, but we should still be concerned, and perhaps very much so.

The use of state government to punish local governments or officials raises concerns distinct from the majoritarian concerns discussed above. The product of the legislature may be majoritarian, but it is undeniably ugly. These concerns, therefore, are akin to the “old-fashioned” concerns about majorities picking on minorities.[319] Here, the minorities may not necessarily be of the traditional protected classes (although in some cases they may be), but rather of a geopolitical stripe. Hence, punitive preemptive acts may raise the kinds of concerns that sound in First Amendment, Equal Protection, and state special legislation frameworks.[320]

What is potentially different is that some of the punitive preemption may not plausibly represent a majority; rather, it results from the actions of minorities that have locked themselves into power through gerrymandering. This makes punitive preemption even more concerning from a democratic perspective than it would be otherwise and should add to the concern currently growing among courts, commentators, and citizens regarding this trend.


While preemption is bad from the perspective of local governance, not all preemption is created equal. Some instances of preemption are more democratically legitimate than others. Assuming that one accepts the concerns articulated herein as legitimate, a variety of potential solutions—some new, some old; some radical, some minor—to democratically suspect preemption are available. If a gerrymandered state legislature is the predominant concern, this can be tackled directly through districting reform. Indeed, there is significant momentum in that direction of late, with many states’ voters adopting districting reform by ballot.[321] In many other states, however, the chances for districting reform will be slim where an entrenched minority (or majority) controls the legislature and there is no direct democracy available to circumvent it. With the Supreme Court having ruled the question of political gerrymandering nonjusticiable, state constitutional attacks will be the only path available for litigation in these states.[322] When some of these attempts inevitably fail, calling a state constitutional convention would seem to be the only other, if unlikely, path out of the morass.

As noted previously, districting reform by itself will not be a panacea, whether because demographic trends make “unintentional gerrymandering” a powerful force or because old preemption lives on from before such reforms were adopted. In these instances, advocates for local control would likely do better by appealing to traditional “home-rule” arguments regarding local control, where available. With respect to preemption imposed by initiative, potential solutions include many of those previously proposed to the initiative process, such as eliminating it entirely, tempering it through campaign donation limits (currently not permitted by Supreme Court doctrine),[323] limiting the subject matters to which it applies,[324] and improving voter education. With respect to the latter, a special emphasis on educating voters about the effects of an initiative on local power would be useful. With respect to repealing pre-existing preemption, however, the benefits of such a move are likely to be limited, particularly when opponents—already in a better position on the “no” side—have the financial advantage.

In sum, the Hunter power that the Supreme Court has recognized that states possess over their localities is an awesome one, tempered only by self-imposed state constitutional restraints and some key constitutional norms like Equal Protection under the Fourteenth Amendment. To be worthy of exercising this power, state governments must represent the majority of the state’s voters in a credible way. While Reynolds was a key step toward ensuring this, it has proved woefully insufficient. Further reform is necessary. In the absence of such reform, many states will continue to run roughshod over their cities and counties in a way that harms local democracy and undermines the promise of home rule to ameliorate shortcomings in democracy at the state and federal levels.[325]

*   Professor of Law, Willamette University College of Law, Salem, Oregon. J.D., University of Michigan. B.S., B.A.S., University of Pennsylvania. The author thanks Jessica Bulman-Pozen, Nestor Davidson, Christopher Elmendorf, Richard Schragger, Ken Stahl, and many others for insightful comments on earlier drafts. The author also thanks Mary Rumsey, Research Librarian at Willamette, and Mick Harris, law student, for outstanding research assistance.

        [1].    Erin Adele Scharff, Hyper Preemption: A Re-ordering of the State-Local Relationship?, 106 Geo. L.J. 1469 (2018). Another commentator has called this phenomenon “super preemption,” Bradley Pough, Understanding the Rise of Super Preemption in State Legislatures, 34 J.L. & Pol. 67 (2018), while Professor Richard Briffault calls it the “new preemption.” Richard Briffault, The Challenge of the New Preemption, 70 Stan. L. Rev. 1995 (2018). For more on the phenomenon, irrespective of nomenclature, see Richard Briffault et al., Am. Const. Soc’y, The Troubling Turn in State Preemption: The Assault on Progressive Cities and How Cities Can Respond (2017) [hereinafter ACS Issue Brief], []; Richard C. Schragger, The Attack on American Cities, 96 Tex. L. Rev. 1163 (2018).

        [2].    See ACS Issue Brief, supra note 1, at 2.

        [3].    See Ariana R. Levinson et al., Federal Preemption of Local Right-to-Work Ordinances, 54 Harv. J. on Legis. 457, 459–60, 491 (2017) (discussing advocacy by “conservative interest groups” for local right-to-work ordinances).

        [4].    See, e.g., Lynn A. Baker & Daniel B. Rodriguez, Constitutional Home Rule and Judicial Scrutiny, 86 Denv. U. L. Rev. 1337, 1370–72 (2009) (endorsing judicial “attempts to police the boundaries between cities and states” as “important”).

        [5].    Justice Hans Linde’s condemnation of the state/local division as a member of the Oregon Supreme Court was particularly sharp:

Nor is it generally useful to define a “subject” of legislation and assign it to one or the other level of government. . . . A search for a predominant state or local interest in the “subject matter” of legislation can only substitute for the political process . . . the court’s own political judgment whether the state or the local policy should prevail.

City of La Grande v. Pub. Emps. Ret. Bd., 576 P.2d 1204, 1213–14 (Or.), aff’d on reh’g, 586 P.2d 765 (Or. 1978).

        [6].    Nestor M. Davidson, The Dilemma of Localism in an Era of Polarization, 128 Yale L.J. 954, 990 (2019).

        [7].    By “deregulatory preemption,” I mean the increasingly common move by state legislatures to repeal city authority over a broad subject area without enacting any statewide regulatory regime in that area. Schragger, supra note 1, at 1182 (attributing this phrase to Professor Richard Briffault).

        [8].    For guidance in that inquiry, Davidson suggests looking to the individual-rights provisions of a state’s constitution, which “can provide insight into the normative commitments of a given state.” Davidson, supra note 6, at 987.

        [9].    207 U.S. 161, 178–79 (1907).

      [10].    Id. at 178; see also Aaron Saiger, Local Government as a Choice of Agency Form, 77 Ohio St. L.J. 423 (2016).

      [11].    377 U.S. 533, 565 (1964) (“[I]n a society ostensibly grounded on representative government, it would seem reasonable that a majority of the people of a State could elect a majority of that State’s legislators.”).

      [12].    See supra note 7 and accompanying text (discussing “deregulatory preemption”).

      [13].    For more on special legislation and how state constitutions deal with that species, see Justin R. Long, State Constitutional Prohibitions on Special Laws, 60 Clev. St. L. Rev. 719 (2012), and Anthony Schutz, State Constitutional Restrictions on Special Legislation as Structural Restraints, 40 J. Legis. 39 (2014).

      [14].    Louise Weinberg, Fear and Federalism, 23 Ohio N.U. L. Rev. 1295, 1304–05 (1997) (noting that under the “compact theory” of state sovereignty, “the states preceded the Union, delegated only a portion of their preexisting powers to the Union, and reserved the rest exclusively”). The Supreme Court rejected this view in Texas v. White, in which it held that Texas could not unilaterally secede from the Union and acts undertaken by Texas in a state of rebellion were nullities. 74 U.S. (7 Wall.) 700–26 (1869).

      [15].    See Franchise Tax Bd. v. Hyatt, 139 S. Ct. 1485, 1493 (2019) (discussing sovereign status of states after Independence); Nat’l Fed’n of Indep. Bus. v. Sebelius, 567 U.S. 519, 535–36 (2012) (contrasting state governments, which need no express authorization to exercise a particular power, with the more limited powers of the federal government); Alden v. Maine, 527 U.S. 706, 713 (1999) (“States’ immunity from suit is a fundamental aspect of the sovereignty which the States enjoyed before the ratification of the Constitution, and which they retain today . . . .”).

      [16].    W.F. Dodd, The Function of a State Constitution, 30 Pol. Sci. Q. 201, 201, 205 (1915) (“[S]tate legislatures inherited the powers of the British parliament and possess such powers in full unless denied.”).

      [17].    See Edmund S. Morgan, Inventing the People 18 (1988) (“In England in the first half of the seventeenth century the doctrine of the divine right of kings . . . reached a culmination.”).

      [18].    Id. at 56 (“A new ideology, a new rationale, a new set of fictions was necessary to justify a government [in 1640s England] in which the authority of kings stood below that of the people or their representatives.”).

      [19].    Id. at 239–87.

      [20].    Donald Ratcliffe, The Right To Vote and the Rise of Democracy, 1787–1828, 33 J. Early Republic 219, 223–30 (2013) (discussing the voting qualifications of the various states after the Revolution). The rules within states were often dizzyingly complicated and subject to change. In Massachusetts, for instance, all free white males over twenty-one could vote in a 1776 election to draft the new state’s inaugural constitution. Chilton Williamson, American Suffrage: From Property to Democracy, 1760–1860, at 101 (1960). The 1780 constitution that was ultimately adopted established a sixty-pound property qualification for all state elections. Id. at 102. An act of 1782 established a separate standard for town elections: real property ownership, payment of a poll tax, or payment of a tax equal to two-thirds of a poll tax. Id. at 103–04.

      [21].    Jill Lepore, Rock, Paper, Scissors: How We Used To Vote, New Yorker (Oct. 6, 2008), [https://perm]. But see Ratcliffe, supra note 20, at 229–30 (arguing that the percentage may have been higher). Since franchise requirements varied even within states based on the specific type of election, it is difficult to come up with a precise figure.

      [22].    Morgan, supra note 17, at 146–47.

      [23].    Id. at 145–46.

      [24].    Id. at 146.

      [25].    Robert G. Dixon, Jr., Democratic Represenation: Restatement in Law and Politics 60–61 (1968); Douglas Keith & Eric Petry, Brennan Ctr. for Justice, Apportionment of State Legislatures, 1776–1920, at 3 (2015), https://www.brennancen (“For the first three decades after 1776, a majority of state constitutions adopted apportioned legislative districts using geography as a basis . . . .”) [].

      [26].    Morgan, supra note 17, at 68.

      [27].    U.S. Const. art. I, § 2, cl. 3 (“Representatives . . . shall be apportioned among the several States . . . according to their respective Numbers . . . .”).

      [28].    See Keith & Petry, supra note 25, at 3–4.

      [29].    Id. at 4.

      [30].    Id. at 1.

      [31].    Stephen Ansolabehere & Samuel Issacharoff, The Story of Baker v. Carr, in Constitutional Law Stories 247, 298–99 (Michael C. Dorf ed., 2004) (noting that rural state legislators in many states resisted reapportionment, “decade after decade defying the mandate of their state constitution”).

      [32].    Reynolds v. Sims, 377 U.S. 533, 545–46 (1964). Tennessee, whose apportionment system was challenged in Baker v. Carr, had similarly failed to reapportion after 1900. 369 U.S. 186, 192–93 (1962).

      [33].    See, e.g., Ansolabehere & Issacharoff, supra note 31, at 298 (“As urban areas grew, the malapportionment of representatives increased.”); John P. White & Norman C. Thomas, Urban and Rural Representation and State Legislative Apportionment, 17 W. Pol. Q. 724, 729–30 (1964) (measuring the widespread overrepresentation of rural populations in state legislatures in the early 1960s).

      [34].    Ward E.Y. Elliott, The Rise of Guardian Democracy 13–16 (1974) (“[U]rban political lobbies, such as the National Municipal League and the United States Council of Mayors” were the main backers of reapportionment before 1954; after 1954 they were joined by “a succession of skilled professional scholars and consultants, backed by the most prestigious foundations and institutions.”).

      [35].    369 U.S. 186, 209 (1962).

      [36].    377 U.S. 533, 568 (1964).

      [37].    Wesberry v. Sanders, 376 U.S. 1, 7–8 (1964).

      [38].    Keith & Petry, supra note 25, at 8.

      [39].    Id. at 6 (discussing this practice in Massachusetts, New Hampshire, and North Carolina).

      [40].    Id. at 4–5.

      [41].    U.S. Const. amend. XIX.

      [42].    Pub. L. No. 89-110, 79 Stat. 437 (1965) (codified as amended at 52 U.S.C. §§ 10101, 10301, 10701). The Fifteenth Amendment to the Constitution, passed after the Civil War in 1870, attempted to prohibit the denial or abridgement of the right to vote “on account of race, color, or previous condition of servitude.” U.S. Const. amend. XV. After Reconstruction, however, states in the South—and elsewhere—routinely flouted the Amendment, which was only weakly enforced by the federal government and courts for years. See Chandler Davidson, The Voting Rights Act: A Brief History, in Controversies in Minority Voting: The Voting Rights Act in Perspective (Bernard Grofman & Chandler Davidson eds., 1992).

      [43].    Keith & Petry, supra note 25, at 6–7.

      [44].    Evenwel v. Abbott, 136 S. Ct. 1120, 1123 (2016).

      [45].    Michael Wines, How the Supreme Court’s Decision on the Census Could Alter American Politics, N.Y. Times (Apr. 23, 2019), ens-census-political-maps.html (discussing efforts in Arizona, Missouri, Nebraska, and Texas) [].

      [46].    Dep’t of Commerce v. New York, 139 S. Ct. 2551, 2576 (2019) (upholding the district court’s remanding of the citizenship-data-collection rule to the Commerce Department, in which the Census Bureau is housed, for further explanation of the agency action).

      [47].    See Exec. Order No. 13,880, § 1, 84 Fed. Reg. 33,821 (July 11, 2019) (recognizing that there was “no practical mechanism for including the [citizenship] question on the 2020 decennial census”).

      [48].    Id. at 33,821, 33,823 (“Nevertheless, we shall ensure that accurate citizenship data is compiled in connection with the census by other means. . . . [I]t may be open to States to design State and local legislative districts based on the population of voter-eligible citizens.”).

      [49].    E.g., People ex rel. Le Roy v. Hurlbut, 24 Mich. 44, 108 (1871) (Cooley, J., concurring) (“[L]ocal government is matter of absolute right; and the state cannot take it away.”); Amasa M. Eaton, The Right to Local Self-Government, 13 Harv. L. Rev. 441, 447 (1900) (arguing that because towns’ lawmaking powers chronologically preceded the colonies’ (later states’) in New England, New York, and some other states, it can’t be said that state constitutions vested all legislative power in state legislatures without reserving towns’ powers over “local affairs”).

      [50].    Hunter v. Pittsburgh, 207 U.S. 161, 178 (1907) (“The number, nature and duration of the powers conferred upon [municipal] corporations and the territory over which they shall be exercised rests in the absolute discretion of the State.”); see also Holt Civic Club v. City of Tuscaloosa, 439 U.S. 60, 71 (1978) (noting “the extraordinarily wide latitude that States have in creating various types of political subdivisions and conferring authority upon them”).

      [51].    The Court has recognized some limits on the breadth of the state’s power over its subentities. For instance, the Court has held that it is unconstitutional for a state legislature to redraw a city’s boundaries for racially discriminatory reasons. See Gomillion v. Lightfoot, 364 U.S. 339 (1960) (invalidating Alabama law that redrew the boundaries of Tuskegee so as to remove the vast majority of black residents from the city).

      [52].    Reynolds v. Sims, 377 U.S. 533, 575 (1964).

      [53].    Id.

      [54].    390 U.S. 474, 481 (1968).

      [55].    Id. (“[T]he States universally leave much policy and decisionmaking to their governmental subdivisions.”).

      [56].    E.g., Hadley v. Junior Coll. Dist., 397 U.S. 50, 66 (1970) (Harlan, J., dissenting) (“[T]here is a much smaller danger of abuse through malapportionment in the case of local units because there exist avenues of political redress that are not similarly available to correct malapportionment of state legislatures.”); Avery, 390 U.S. at 489 (Harlan, J., dissenting) (“Local governments are creatures of the States,  and  they  may  be  reformed . . . by the state legislatures, which are now required to be apportioned according to Reynolds . . . .”).

      [57].    Ball v. James, 451 U.S. 355, 373 (1981) (Powell, J., concurring).

      [58].    E.g., Jon Elster, Nuts and Bolts for the Social Sciences 155 (1989) (“[T]he notion of a popular will is incoherent, or . . . the popular will is itself incoherent, whichever you prefer.”).

      [59].    Baker & Rodriguez, supra note 4, at 1343, app. at 1374–1424.

      [60].    See Paul Diller, Intrastate Preemption, 87 B.U. L. Rev. 1113, 1126–27 (2007) (noting the wide berth state legislatures have to preempt in most states).

      [61].    Reynolds v. Sims, 377 U.S. 533, 565 (1964) (“Full and effective participation by all citizens in state government requires, therefore, that each citizen have an equally effective voice in the election of members of his state legislature.”).

      [62].    Stephen Ansolabehere & James M. Snyder, Jr., The End of Inequality: One Person, One Vote and the Transformation of American Politics 199 (2008) (observing that urban, previously underrepresented Florida counties saw a marked increase in their share of state funding between 1960 and 1980). But see Roy A. Schotland, Commentary, The Limits of Being “Present at the Creation,” 80 N.C. L. Rev. 1505, 1505 (2002) (quoting California lawmaker Jesse Unruh as asserting that cities fared better before one-person, one-vote because they could “work things out with the agricultural areas,” whereas after one-person, one-vote, suburbs favoring low taxes were empowered).

      [63].    See generally Jonathan A. Rodden, Why Cities Lose (2019).

      [64].    Id. at 88 (noting that the two major political parties “went from having incoherent or indistinguishable platforms on social issues to taking clear and divergent positions on these salient issues”); see also Daniel J. Hopkins, The Increasingly United States (2018) (arguing that state and local politics have become markedly more nationalized since the 1960s).

      [65].    Rodden, supra note 63, at 7 (“For several decades, self-styled Democratic candidates were able to distance themselves from their party’s presidential candidates and win in Republican-leaning districts.”).

      [66].    Id. at 8 (“[T]he transformation of the Democrats into a truly urban party was delayed by their long-lasting coalition with rural Southern segregationists, which began to fray during the civil rights era and was only fully severed in the 1990s.”); see also Eric Schickler, Disjointed Pluralism: Institutional Innovation and Development of the U.S. Congress 163–74 (2001) (highlighting the influence of conservative Southern Democrats on Congressional decision making, particularly through the House Rules Committee).

      [67].    So named for their association with the moderate views of former New York governor and later vice president Nelson Rockefeller. See generally Richard Norton Smith, On His Own Terms: A Life of Nelson Rockefeller (2014). The decline of urban liberal Republicans began in the 1930s and continued through the 1960s. See Thomas K. Ogorzalek, The Cities on the Hill: How Urban Institutions Transform National Politics 209, 214 fig.6.4 (2018).

      [68].    See Vincent J. Cannato, The Ungovernable City: John Lindsay and His Struggle to Save New York 7–18 (2001). Lindsay’s district also included Lower East Side tenements and parts of (then-)bohemian Greenwich Village. Id. at 5–6.

      [69].    Cf. Seth J. Hill & Chris Tausanovitch, Southern Realignment, Party Sorting, and the Polarization of the American Primary Electorates, 1958–2012, 176 Pub. Choice 107, 115 (2018) (“In the 1950s, the [political ideology of the] median Democratic primary voter was indistinguishable from the median voter in the public as a whole . . . .”).

      [70].    See Jessica Bulman-Pozen, Partisan Federalism, 127 Harv. L. Rev. 1077, 1086–87 & nn.24–30 (2014) (arguing that today’s state and national parties are more partisan and ideologically cohesive than they were decades earlier and citing numerous sources to support that point).

      [71].    Keith T. Poole & Howard Rosenthal, Ideology & Congress 318 (2007) (“The collapse of the old southern Democratic Party has produced . . . two sharply distinct political parties,” reflecting a “degree of polarization in Congress . . . approaching levels not seen since the 1890s”).

      [72].    See Cliff Schecter, Extremely Motivated: The Republican Party’s March to the Right, 29 Fordham Urb. L.J. 1663 (2002). On immigration, see Don Gonyea, The GOP’s Evolution on Immigration, NPR Politics (Jan. 25, 2018), the-gops-evolution-on-immigration [].

      [73].    See Lazaro Gamio, Urban and Rural America Are Becoming Increasingly Polarized, Wash. Post (Nov. 17, 2016), ion/urban-rural-vote-swing/ [].

      [74].    Jowei Chen & Jonathan Rodden, Unintentional Gerrymandering: Political Geography and Electoral Bias in Legislatures, 8 Q.J. Pol. Sci. 239, 241, 264 (2013).

      [75].    See David Daley, Ratf**ked: The True Story Behind the Secret Plan To Steal America’s Democracy (2016) (detailing intentional gerrymandering, particularly by Republicans, after 2010); Jowei Chen, The Impact of Political Geography on Wisconsin Redistricting: An Analysis of Wisconsin’s Act 43 Assembly Districting Plan, 16 Election L.J. 1, 10 (2017) (concluding that the allegedly intentionally gerrymandered Wisconsin state assembly district map “created an extremely biased Assembly plan” that favored Republicans far more than 200 other simulated maps drawn using more neutral districting criteria).

      [76].    138 S. Ct. 1916, 1923 (2018). For easy access to the litigation documents in Gill, see Gill v. Whitford, Documents, Campaign Legal Ctr., ns/gill-v-whitford []. The case was originally denominated Whitford v. Nichol, so plaintiffs’ expert reports are listed as “Nichol Expert Reports.”

      [77].    Gill, 138 S. Ct. at 1923–24.

      [78].    Id. at 1933–34 (concluding that the plaintiffs had “fail[ed] to demonstrate Article III standing”).

      [79].    139 S. Ct. 2484, 2506–07 (2019).

      [80].    Plaintiffs’ Opposition to Defendant’s Motion for Summary Judgment at 1, Whitford v. Nichol, 180 F. Supp. 3d 583 (W.D. Wisc. 2016) (No. 15-cv-421-bbc).

      [81].    See id. at 36–40 (discussing the “efficiency gap”); see also Eric McGhee, Measuring Partisan Bias in Single-Member District Electoral Systems, 39 Legis. Stud. Q. 55 (2014); Nicholas O. Stephanopoulos & Eric M. McGhee, Partisan Gerrymandering and the Efficiency Gap, 82 U. Chi. L. Rev. 831, 831 (2015).

      [82].    See Expert Report of Siman Jackman, Assessing the Current Wisconsin State Legislative Redistricting Plan at 19–36, Whitford, 180 F. Supp. 3d 583 (No. 15-cv-421-bbc) [hereinafter Jackman Report], FORD%20V.%20NICHOL-Report_0.pdf (analyzing the efficiency gap in state legislative elections from 1972 to 2014) [].

      [83].    See supra text accompanying notes 70–72.

      [84].    David A. Leib, Analysis Indicates Partisan Gerrymandering Has Benefited GOP, AP (June 25, 2017), [https://per]; see also AP, Measuring the Effects of Gerrymandering, [hereinafter AP 2016 Chart] (chart of efficiency gap in 2016 state house races) [].

      [85].    The only (partial) exceptions are Mississippi and Vermont. Mississippi uses an “electoral” system similar to that used to elect the President, with state house districts serving the role of states in the federal system. Miss. Const. art. V, § 140. When no candidate receives both an electoral and popular majority, the house of representatives picks the governor by viva voce vote. Id. § 141. In Vermont, members of the legislature from both houses, voting jointly, select the governor by secret ballot when no candidate receives an outright majority. Vt. Const. ch. II, § 47; see also D. Gregory Sanford & Paul Gillies, And if There Be No Choice Made: A Meditation on Section 47 of the Vermont Constitution, 27 Vt. L. Rev. 783, 783 (2003).

      [86].    Cf. McNollgast, Positive Canons: The Role of Legislative Bargains in Statutory Interpretation, 80 Geo. L.J. 705, 720 (1992) (using this term to describe kill points for legislation).

      [87].    To be sure, in some instances there are governors elected with relatively small pluralities, such as when there is a third-party candidate that attracts a significant portion of the vote. See Jack Santucci, Maine Ranked-Choice Voting as a Source of Electoral-System Change, 54 Representation 297, 297–98 (2018) (discussing “widespread dissatisfaction” with Maine’s two-term governor, Paul LePage, who won 38% in his first winning race (2010) and 48% in his second winning race (2014)).

      [88].    See generally Miriam Seifter, Gubernatorial Administration, 131 Harv. L. Rev. 483 (2017).

      [89].    See Nat’l Conference of State Legislatures, Inside the Legislative Process, tbl.98-6.22, (2008 rev., 2009), (listing five states as requiring mere majorities to override vetoes of at least some bills and seven that require three-fifths majorities to override vetoes) [].

      [90].    Cf. Keith T. Poole & Howard Rosenthal, The Polarization of American Politics, 46 J. Pol. 1061, 1064–66, 1071–72 (1984) (noting that U.S. senators from the same state but from different parties have highly dissimilar voting records, suggesting that each party represents an extreme support coalition in the state).

      [91].    See infra Sections II.B.3, II.B.5.

      [92].    See supra note 47 and accompanying text.

      [93].    Diller, supra note 60, at 1124–27 (describing the emergence of municipal home rule among the states).

      [94].    See sources cited supra note 1.

      [95].    Cf. Justin McCarthy, Americans Still More Trusting of Local than State Government, (Oct. 8, 2018), (explaining that 72% of U.S. adults trust their local government “a great deal” or “fair amount” while only 63% of U.S. adults “say the same about their state government”) [].

      [96].    See, e.g., Schragger, supra note 1, at 1166–67.

      [97].    See infra Section II.B.

      [98].    There is, of course, a substantial volume of literature on the role interest groups and campaign finance play in the political process. For some of the classics on interest groups, see Mancur Olson, Jr., The Logic of Collective Action: Public Goods and the Theory of Groups (1965), and Russell Hardin, Collective Action (1982). On campaign finance, see Gene M. Grossman & Elhanan Helpman, Special Interest Politics (2001), and Lynda W. Powell, The Influence of Campaign Contributions in State Legislatures (2012).

      [99].    See Alexander Hertel-Fernandez, State Capture 238–41 (2019) (discussing ALEC’s strategy of seeking state legislative preemption of local minimum wage and paid sick-leave ordinances); Schragger, supra note 1, at 1170 (describing ALEC’s role in lobbying for preemption in state legislatures).

    [100].    Rhode Island’s Democratic-dominated legislature, for instance, preempted local minimum wage authority in 2014, 2014 R.I. Pub. Laws, ch. 145, art. 11, § 4 (codified at R.I. Gen. Laws § 28-12-25), stifling Providence’s effort to raise the minimum wage for certain workers.  Sam Adler-Bell, Why Are Rhode Island Democrats Blocking Minimum-Wage Increases?, Nation (June 11, 2014), (noting that Providence was on the verge of passing a $15/hour minimum wage law for employees of large hotels) [ 46GG-7XUS].

    [101].    See Diller, supra note 60, at 1136 (describing public-sector unions’ push for preemption in certain areas). Public-sector unions’ political power, of course, may now be reduced after the Supreme Court’s decision limiting “fair share” or “compulsory” union dues in Janus v. AFSCME Council 31, 138 S. Ct. 2448, 2464, 2487 (2018).

    [102].    See Daley, supra note 75, at 4–6 (explaining the millions of dollars invested in 2010 campaigns by large American corporations and interest groups in the hopes of influencing the drawing of district maps afterwards).

    [103].    Whitford v. Gill, 218 F. Supp. 3d 837, 903–04 (W.D. Wis. 2016), rev’d, 138 S. Ct. 1916, 1923–24 (2018).

    [104].    Id. at 907.

    [105].    Id. at 860–61.

    [106].    Id. at 910.

    [107].    These examples closely track those that were highlighted in an amicus brief co-authored by myself that was submitted to the Supreme Court in Gill v. Whitford. Brief of International Municipal Lawyers Association et al., as Amici Curiae Supporting Appellees at 16, 20, Gill v. Whitford, 138 S. Ct. 1916 (2018) (No. 16-1161). Those examples did not include 2018 data and some additional information offered here.

    [108].    See, e.g., Bush v. Gore, 531 U.S. 98 (2000).

    [109].    See United States Presidential Election Results, David Leip’s Atlas U.S. Presidential Elections, [hereinafter Leip], (click to presidential election results in 2000, 2004, 2008, 2012, and 2016, and hover over Florida for each)  [].

    [110].    Id. (click to Florida gubernatorial election results for 2018, 2014, and 2010).

    [111].    Rhodes Cook, America Votes 31: Election Returns by State 2013–2014, at 81 (2015); Florida Gubernatorial and Lieutenant Gubernatorial Election, 2018,  Ballotpedia,,_ 2018 [].

    [112].    Leip, supra note 109 (click to Florida Senate election results for 2018).

    [113].    See Florida State Legislature, Ballotpedia, Legislature [].

    [114].    Florida House of Representatives Elections, 2014, Ballotpedia, https://ballotped,_2014 []; Florida State Senate Elections, 2014, Ballotpedia, Senate_elections,_2014 [].

    [115].    Martinez v. Bush, 234 F. Supp. 2d 1275, 1296 (S.D. Fla. 2002).

    [116].    Jackman Report, supra note 82, at 73 fig.36.

    [117].    AP 2016 Chart, supra note 84.

    [118].    Cf. Chen & Rodden, supra note 74, at 244 (noting that “Democrats in Florida [are] highly concentrated in downtown Miami” and several other large cities, while suburbs and rural areas “are generally Republican”). It should be noted that the Miami metropolitan area’s sizable Cuban-American population has traditionally voted Republican and, hence, many Republican state legislators have hailed from the Miami area with strong Cuban-American support. See, e.g., The Cuban Vote in Florida, NPR: Weekend Edition Sunday  (Nov. 11, 2018), [].

    [119].    See Cook, supra note 111, at 80.

    [120].    See Russell Berman, Florida Struggles To Pay the Tab for Rejecting Obamacare, Atlantic (May 8, 2015), gles-to-pay-the-tab-for-rejecting-obamacare/392678/ (discussing the Florida house’s resistance to Medicaid expansion despite state senate and sometimes gubernatorial support for the proposal) [].

    [121].    Act of June 14, 2013, ch. 2013-200, sec. 1, 2013 Fla. Laws 2508 (codified as amended at Fla. Stat. § 218.077).

    [122].    City of Miami Beach v. Fla. Retail Fed’n, Inc., 233 So. 3d 1236, 1240 (Fla. Dist. Ct. App. 2017).

    [123].    Fla. Min. Wage Amendment, Amendment 5 (2004) (codified at Fla. Const. art. X, § 23).

    [124].    City of Miami Beach, 233 So. 3d at 1238.

    [125].    Act of March 16, 2016, ch. 2016-61, sec. 7, 2016 Fla. Laws 519, 525 (codified at Fla. Stat. § 500.90).

    [126].    Fla. Stat. § 500.90 (exempting “local ordinances or provisions thereof enacted before January 1, 2016”); Fla. Retail Fed’n, Inc. v. City of Coral Gables, No. 2016-018370-CA-01, slip op. at 6–10 (Fla. Cir. Ct. Mar. 9, 2017) (invalidating state preemption as to cities within Miami-Dade County, which enjoy constitutional home-rule immunity to special legislation), rev’d, No. 3D17-562, 2019 Fla. App. LEXIS 12501, at *2 (Fla. Dist. Ct. App. Aug. 14, 2019).

    [127].    Act of June 2, 2011, ch. 2011-109, sec. 1, 2011 Fla. Laws 1767, 1768‒70 (codified as amended at Fla. Stat. § 790.33).

    [128].    E.g., Fla. Carry, Inc. v. City of Tallahassee, 212 So. 3d 452, 455–56 (Fla. Dist. Ct. App. 2017). In a recent decision, a Florida circuit court invalidated these penalty provisions using a variety of theories, including separation of powers. City of Weston v. DeSantis, No. 2018 CA 0699 (Fla. Cir. Ct., July 26, 2019).

    [129].    S.B. 168, 2019 Leg. Sess. (Fla. 2019) (to be codified at Fla. Stat. §§ 908.101–.109).

    [130].    Julia Ingram, South Miami Sues To Block Sanctuary City Ban, Says It Will Divide Police and Residents, Miami Herald (July 16, 2019), local/immigration/article232713882.html (noting that South Miami was the only municipality to sue Florida over the law) [].

    [131].    See supra notes 103 and 104 and accompanying text.

    [132].    Florida House of Representatives Elections, 2018, Ballotpedia, https://ballotped,_2018 [].

    [133].    Florida redrew its state senate maps in 2015 after they were challenged as improperly gerrymandered under the Fair Districts Amendment to the state constitution, passed by the voters in 2010. In re Senate Joint Research on Legislative Apportionment 1176, 83 So. 3d 597, 598 (Fla. 2012). The new senate map, which was ultimately ordered by a court as a settlement to subsequent litigation, took effect before the 2016 elections. See League of Women Voters v. Detzner, 179 So. 3d 258, 260 (Fla. 2015). No one has challenged the Florida state house maps. Under the Fair Districts Amendment, district lines are only struck down if someone brings a lawsuit challenging them and can prove political intent in drawing them. See Fla. Const. art. III, § 21 (“No apportionment plan or district shall be drawn with the intent to favor or disfavor a political party or an incumbent . . . .”).

    [134].    Detzner, 179 So. 3d at 260.

    [135].    Florida State Legislature, Ballotpedia, Leg islature [].

    [136].    Id.

    [137].    Leip, supra note 109 (click to 2016 presidential election results and hover over Michigan).

    [138].    Lauren Gibbons, How Michigan’s Presidential Election Map Has Changed Since the 1980s, (Oct. 27, 2016), aways_from_three_deca.html [].

    [139].    Cook, supra note 111, at 188; David Eggert, Democrat Whitmer Wins Michigan Governor Race, Beats Schuette, AP News (Nov. 7, 2018), fa4319b6f6ccd63243b9d5 []; Michigan, States in the Senate, [].

    [140].    Michigan House of Representatives Elections, 2012, Ballotpedia, https://ballotped,_2012 [ K8L].

    [141].    Michigan House of Representatives Elections, 2014, Ballotpedia, https://ballotped,_2014 [ 3YY].

    [142].    Jackman Report, supra note 82, at 73 fig.36.

    [143].    David A. Leib, AP Analysis Indicates Partisan Gerrymandering Has Benefited GOP, AP News (June 25, 2017), (“Last fall, voters statewide split their ballots essentially 50-50 . . . [y]et Republicans won 57 percent of the [state] House seats, claiming 63 seats to the Democrats’ 47.”) [https: //].

    [144].    AP 2016 Chart, supra note 84.

    [145].    Jordan Newton, A Midterm Retrospective: The 2018 Election and Gerrymandering, Citizens Res. Council Mich. (Nov. 15, 2018), [].

    [146].    Id.

    [147].    Jackman Report, supra note 82, at 71–73.

    [148].    Michigan, All About Redistricting, [].

    [149].    Local Financial Stability and Choice Act, No. 436, 2012 Mich. Pub. Acts 1887 (codified at Mich. Comp. Laws §§ 141.1541–.1575); Local Government and School District Fiscal Accountability Act, No. 4, 2011 Mich. Pub. Acts 14 (rejected via referendum as Proposal 1 in 2012); see also In re City of Detroit, 504 B.R. 97, 121–22 (Bankr. E.D. Mich. 2013) (reviewing history of the emergency manager law used to steer Detroit into bankruptcy).

    [150].    Paul Egan, Is Emergency Manager Law to Blame for Flint Water Crisis?, Detroit Free Press (Oct. 24, 2015), cy-manager-law-blame-flint-water-crisis/74048854/ (last updated Oct. 25, 2015) [https://].

    [151].    Local Government Labor Regulatory Agency Act, No. 105, 2015 Mich. Pub. Acts 653 (codified at Mich. Comp. Laws §§ 123.1381–.1396); see also Emily Lawler, Gov. Rick Snyder Signs ‘Death Star’ Bill Prohibiting Local Wage, Benefits Ordinances, (June 30, 2015), _st.html) (last updated Jan. 20, 2019) [].

    [152].    Act of Dec. 28, 2016, No. 389, 2016 Mich. Pub. Acts 1779 (codified at Mich. Comp. Laws §§ 445.591–.593).

    [153].    See Washtenaw County, Mich., Carryout Bag Ordinance, https://localannarbor. [].

    [154].    See Emily Lawler, Ban on Local Plastic Bag Bans Now Michigan Law, (Dec. 19, 2016), bans.html (last updated Jan. 19, 2019) [].

    [155].    Chris Kardish, How North Carolina Turned So Red So Fast, Governing (July 2014), m.html [].

    [156].    Cook, supra note 111, at 267.

    [157].    Leip, supra note 109 (click to 2012 and 2016 presidential election results and hover over North Carolina).

    [158].    See H.B. 937, 2011 Gen. Assemb., Reg. Sess. (N.C. 2011); House Roll Call Vote Transcript for Roll Call #1287, North Carolina General Assembly, https://www.ncleg. gov/Legislation/Votes/RollCallVoteTranscript/2011/H/1287 [].

    [159].    Jackman Report, supra note 82, at 73 fig.36.

    [160].    AP 2016 Chart, supra note 84.

    [161].    316 F.R.D. 117, 117 (M.D.N.C. 2016).

    [162].    North Carolina v. Covington, 138 S. Ct. 2548, 2550–54 (2018).

    [163].    North Carolina State Senate Elections, 2018, Ballotpedia, https://ballotpedia. org/North_Carolina_State_Senate_elections,_2018 [].

    [164].    North Carolina House of Representatives Elections, 2018, Ballotpedia, https://ball,_2018 [ WZU9-X7Q8].

    [165].    N.C. Const. art. II, § 22.

    [166].    Jason DeBruyn, Dems Win More Votes; Reps Win More Seats, (Nov. 9, 2018), (showing that despite winning 49% of the total statewide vote for senate compared to Democrats’ 51%, Republicans would control 55% of the seats; for the house, Republicans won 49.6% of the vote to Democrats’ 50.4%, but would control 58% of the seats) [].

    [167].    See Frayda S. Bluestein, Do North Carolina Local Governments Need Home Rule?, 84 N.C. L. Rev. 1983, 1985 (2006) (“North Carolina local government powers are established through specific statutory delegations . . . .”); see also Baker & Rodriguez, supra note 4, at 1338 & n.10 (describing North Carolina as having “no [constitutional] home rule at all”).

    [168].    Charlotte, N.C., Code of Ordinances §§ 12-56 to 12-58 (2016) (noting that it was preempted by state law and invalidated by Act of Mar. 23, 2016, S.L. 2016-3, 2016 N.C. Sess. Laws 12).

    [169].    Editorial, North Carolina’s Bigoted Bathroom Law, N.Y. Times, Mar. 25, 2016, at A24.

    [170].    See  Ryan  Bort,  A  Comprehensive  Timeline  of  Public  Figures  Boycotting  North Carolina over the HB2 ‘Bathroom Bill’, Newsweek (Sept. 14, 2016, 5:06 PM), http://www. [].

    [171].    S.L. 2016-3, sec. 2.1, 2016 N.C. Sess. Laws at 14 (codified at N.C. Gen. Stat. § 95-25.1).

    [172].    Act of Mar. 30, 2017, S.L. 2017-4, sec. 1, 2017 N.C. Sess. Laws 81 (codified as amended at N.C. Gen. Stat. § 143-760). Unsatisfied with the partial repeal, transgender advocates continued their lawsuit over HB2. Recently, Governor Roy Cooper agreed with civil rights groups on a settlement that would allow individuals to use the bathroom consistent with their gender identity. Dan Levin, North Carolina Reaches Settlement on ‘Bathroom Bill, N.Y. Times (July 23, 2019), ina-transgender-bathrooms.html [].

    [173].    S.L. 2017-4, 2017 N.C. Sess. Laws 81 (codified as amended at N.C. Gen. Stat. § 143-760).

    [174].    Kevin Robillard, North Carolina Governor Signs Laws Restricting Successor’s Power, Politico (Dec. 16, 2016, 3:41 PM), (discussing legislation that: “reduce[d] the number of positions the governor can hire and fire at will from 1,500 to 300, strip[ped] the governor’s party of the power to control the state board of elections, require[d] legislative approval of gubernatorial cabinet appointments, and move[d] the power to appoint trustees for the University of North Carolina to the legislature”) [].

    [175].    Wendy R. Weiser & Daniel I. Weiner, Brennan Ctr. for Justice, North Carolina Legislature’s Power Grab Disregards Basic Principles of Democracy (Aug. 14, 2018), https: // nciples (discussing proposed initiatives that would limit the governor’s powers over the state’s ethics and elections board and over judicial appointments) []. In the process of proposing these initiatives, the legislature also stripped the power of the pre-existing Constitutional Amendments Publication Commission to draft ballot title language. Id.

    [176].    North Carolina 2018 Ballot Measures, Ballotpedia, _Carolina_2018_ballot_measures (showing the defeats of the Legislative Appointments to Elections Board and Commissions and the Judicial Selection for Midterm Vacancies Amendments) [].

    [177].    See generally Kyle Kondik, The Bellwether—Why Ohio Picks The President (2016).

    [178].    Leip, supra note 109 (click to 2000, 2004, 2008, and 2012 presidential election results and hover over Ohio for each).

    [179].    Id.

    [180].    Governor Kasich was a popular governor and was re-elected in 2014 by a margin of 63.6% to 33%. Cook, supra note 111, at 281.

    [181].    Leip, supra note 109 (click on 2018 Ohio governor’s race).

    [182].    Ohio, States in the Senate, [https: //].

    [183].    Ohio State Senate, Ballotpedia, [https: //].

    [184].    Ohio House of Representatives, Ballotpedia, of_Representatives [].

    [185].    Jackman Report, supra note 82, at 69, 71, 73 fig.36.

    [186].    AP 2016 Chart, supra note 84.

    [187].    Rich Exner, Ohio Democrats Nearly Match Republicans in Statehouse Votes, but Will Remain Deep in the Minority; What’s Ahead for Gerrymandering, (Nov. 30, 2018), crats-outpolled-repub.html [].

    [188].    See David Stebenne, Re-mapping American Politics: The Redistricting Revolution Fifty Years Later, 5 Origins (Feb. 2012), can-politics-redistricting-revolution-fifty-years-later (discussing the geography of gerrymandering in Ohio after the 2010 census) []; see also id. (“What the Republicans tried to do is to create the maximum number of safe Republican seats in the . . . Ohio General Assembly, and a minimum number of truly competitive seats . . . [by] break[ing] up major metropolitan areas (where the Democrats are usually strongest) and combin[ing] pieces of them with exurban, small town and rural areas (where the Republicans are strongest).”).

    [189].    Act of Feb. 22, 2002, No. 386, 2002 Ohio Laws 6938 (codified at Ohio Rev. Code Ann. §§ 1.63, 1349.25–.37).

    [190].    Act of Jan. 30, 2006, No. 82, 2006 Ohio Laws 802 (codified at Ohio Rev. Code Ann. § 9.481), sustained by City of Lima v. State, 909 N.E.2d 616, 618 (Ohio 2009) (rejecting municipal home-rule challenge to state preemption of local residency ordinances).

    [191].    Act of Dec. 13, 2006, No. 347, 2006 Ohio Laws 8138 (codified at Ohio Rev. Code Ann. § 9.68).

    [192].    Act of June 1, 2016, No. 180, 2016 Ohio Laws (codified at Ohio Rev. Code Ann. § 9.75), upheld by City of Cleveland v. State, 2019-ohio-3820 (Ohio Sept. 24, 2019).

    [193].    See Act of Dec. 22, 2014, No. 342, 2014 Ohio Laws (codified at Ohio Rev. Code Ann. §§ 4511, 1901, 1907); see also City of Dayton v. State, 87 N.E.3d 176 (Ohio 2017) (holding that the provisions of the state law that sought to regulate local government use and enforcement of red-light and speed cameras violated the home-rule provision of the Ohio Constitution).

    [194].    See Whitford v. Gill, 218 F. Supp. 3d 837, 902‒10 (W.D. Wis. 2016), vacated, 138 S. Ct. 1916, 1923‒24 (2018).

    [195].    Whitford, 218 F. Supp. 3d at 846–53 (recounting history of Act 43); id. at 896 (“[T]he evidence establishes that one of the purposes of Act 43 was to secure Republican control of the Assembly under any likely future electoral scenario for the remainder of the decade, in other words to entrench the Republican Party in power.”).

    [196].    Id. at 902.

    [197].    Id. at 905 (citing Jackman Report, supra note 82).

    [198].    Id. at 906.  In the Wisconsin assembly, unlike in other states, the EG percentage almost exactly translates into the extra seats won by the party favored by the districting plan in question; this is because the assembly has 99 seats, or almost 100.

    [199].    AP 2016 Chart, supra note 84.

    [200].    David A. Lieb, Election Shows How Gerrymandering Is Difficult to Overcome, U.S. News (Nov. 17, 2018, 3:59 PM), midterm-elections-reveal-effects-of-gerrymandered-districts [ FV].

    [201].    These offices were governor (elected in conjunction with lieutenant governor), attorney general, secretary of state, and treasurer. See Canvass Results for 2018 General Election, Wis. Elections Commission, /Summary%20Results-2018%20Gen%20Election_0.pdf [].

    [202].    See Craig Gilbert, The Red & the Blue: Political Polarization Through the Prism of Metropolitan Milwaukee, Marq. Law., Fall 2014, at 8, 10–11, 14 (demonstrating that the city of Milwaukee leans strongly Democratic while its surrounding suburbs lean strongly Republican); Kazimierz J. Zaniewski & James R. Simmons, Divided Wisconsin: Partisan Spatial Electoral Realignment, 13 Geography Teacher 128, 132 (2016) (noting that Democrats “domina[te] centers in Milwaukee, Madison (and its suburbs), and the state’s medium-sized cities” while Republican majorities have grown in “suburbs, exurbs, small towns, and rural areas”).

    [203].    See Act of June 30, 2013, ch. 20, sec. 1269m, 2013 Wis. Sess. Laws 85, 449 (codified at Wis. Stat. § 66.0418).

    [204].    See Act of April 25, 2016, ch. 374, 2015 Wis. Sess. Laws 1544, 1544‒45 (codified at Wis. Stat. § 66.0438).

    [205].    See ch. 20, sec. 1270, 2013 Wis. Sess. Laws at 449 (codified at Wis. Stat. § 66.0502) (constitutionality upheld in Black v. City of Milwaukee, 882 N.W.2d 333, 338 (Wis. 2016)).

    [206].    Act of June 1, 2005, ch. 12, sec. 1, 2005 Wis. Sess. Laws 23, 24 (codified at Wis. Stat. § 104.001).

    [207].    Act of May 5, 2011, ch. 16, sec. 3, 2011 Wis. Sess. Laws 76, 76–77 (codified at Wis. Stat. § 103.10).

    [208].    Cf. Whitford v. Gill, 218 F. Supp. 3d 837, 901 n.266 (W.D. Wis. 2016) (Act 43’s gerrymandering deprives Democrats of “the opportunity to pass an agenda consistent with their policy objectives”), vacated, 138 S. Ct. 1916 (2018).

    [209].    See Jackman Report, supra note 82, at 73 fig.36; see also AP 2016 chart, supra note 84 (showing that in 2016, Republicans won more seats in the Indiana state house than they would have under more neutral maps).

    [210].    Redistricting in Indiana After the 2010 Census, Ballotpedia, https://ballotpedia. org/Redistricting_in_Indiana_after_the_2010_census [].

    [211].    See Virginia, All About Redistricting, (noting that the state senate had a small Democratic majority in 2011 when state district lines were redrawn, while Republicans controlled the lower house and the governor’s office) []. For more on the effectiveness of Virginia’s gerrymander, even during the initial pro-Democratic “wave” of 2017, see Nicholas Stephanopoulos, Opinion, What Virginia Tells Us, and Doesn’t Tell Us, About Gerrymandering, L.A. Times (Nov. 10, 2017), es-virginia-20171110-story.html [].

    [212].    Mark Joseph Stern, Virginia Democrats’ Victory Proves that Gerrymandering Matters, (Nov. 6, 2019), rats-victory-after-killing-racial-gerrymander.html (asserting that Democrats won more seats in 2019 than in 2017 in part because the prior districting plan was invalidated in the year between as an unconstitutional racial gerrymander) []; see Bethune-Hill v. Va. State Bd. of Elections, 326 F. Supp.2d 128 (E.D. Va. 2018), aff’d, 139 S. Ct. 1945 (2019).

    [213].    New York, All About Redistricting, (noting that the state senate had a slight Republican majority in 2011 when state districts were redrawn, while Democrats controlled the lower house and the governor’s office) []; see also Editorial, Gerrymandering Failed for GOP in State Senate Loss, Buff. News (Nov. 12, 2018), ymandered-corruption []; Joshua Simmons, Partisan Gerrymandering in New York, Bencen Blog (Dec. 1, 2017), ents/2017/12/01/partisan-gerrymandering-in-new-york (“[T]here is an agreement between the Republican-led State Senate and the Democratic-led Assembly that each house majority does their own redistricting and signs off on the other.”) []. In 2018, “[c]hanging demographics and the state’s leftward tilt finally overcame” Republicans’ senate gerrymandering, which had been buttressed by alliances with small groups of “breakaway” Democrats, and Democrats won outright control of the chamber. Gerrymandering Failed for GOP in State Senate Loss, supra; see also Simmons, supra.

    [214].    See Jackman Report, supra note 82, at 73 fig.36; AP 2016 Chart, supra note 84.

    [215].    See Jackman Report, supra note 82, at 73 fig.36; Indeed, Rhode Island’s pro-Demo- cratic EG was 11% in 2012, the fourth highest in the country. Bill Corriher & Liz Kennedy, Distorted Districts, Distorted Laws, Ctr. for Am. Progress (Sept. 19, 2017), https://www. orted-laws/ []. In 2016, however, its pro-Democratic EG was much lower, a mere 2.25%. AP 2016 Chart, supra note 84.

    [216].    See Diller, supra note 60, at 1113, 1148 (2007) (“Legislative inertia is a strong force . . . .”)

    [217].    See Christopher S. Elmendorf, From Educational Adequacy to Representational Adequacy: A New Template for Legal Attacks on Partisan Gerrymanders, 59 Wm. & Mary L. Rev. 1601, 1623–27 (2018) (discussing the possibility of using existing state constitutional text as the basis for partisan gerrymandering claims).

    [218].    In Florida, as noted above, the courts intervened because the voters inserted a clear textual hook requiring nonpolitical districting into the state constitution in 2010.  See supra note 133. The Supreme Court of Pennsylvania and a North Carolina state trial court have recently relied on older, vaguer state constitutional provisions to invalidate partisan gerrymandering.  See League of Women Voters v. Commonwealth, 178 A.3d 737, 802–04 (Pa. 2018) (holding that politically gerrymandered Congressional districts violated the state constitution’s “Free and Equal Elections” clause, Pa. Const. art. I, § 5, the language of which dates back to 1776); Common Cause v. Lewis, No. 18CVS014001, 2019 N.C. Super LEXIS 56, *11 (N.C. Super. Ct. Sept. 3, 2019) (holding that politically drawn state legislative districts violated the state constitution’s “Equal Protection” and “Free Elections” clauses); see also James A. Gardner, Foreword: Representation Without Party: Lessons from State Constitutional Law, 37 Rutgers L.J. 881, 969–70 & n.219 (2006) (discussing the potential role that “free and equal” elections, and equal protection, and other clauses in state constitutions might play in constraining gerrymandering).

    [219].    See Initiative and Referendum States, Nat’l Conf. St. Legislatures, http://www. [https://perm].

    [220].    See supra note 149 and accompanying text.

    [221].    See id.

    [222].    Thomas E. Cronin, Direct Democracy: The Politics of Initiative, Referendum, and Recall 50–51 (2001).

    [223].    Julian N. Eule, Judicial Review of Direct Democracy, 99 Yale L.J. 1503, 1512–13 (1990) (describing the “conventional history” of direct democracy).

    [224].    Kathleen Ferraiolo, State Policy Activism via Direct Democracy in Response to Federal Partisan Polarization, 47 Publius: J. Federalism 378, 378 (2017) (noting that “supporters of marijuana legalization and minimum wage increases have relied on initiatives to compensate for lags in legislatures’ responses to evolving public support for these positions”); id. at 384 (“In the wake of congressional inaction, gun-control groups . . . have turned to state initiative processes on various occasions in the 2010s to enact their favored policies.”).

    [225].    Indeed, on just the shortcomings of the Progressive movement’s creation of direct democracy, see Eule, supra note 223, at 1512 n.38.

    [226].    Hans A. Linde, When Initiative Lawmaking Is Not “Republican Government”: The Campaign Against Homosexuality, 72 Or. L. Rev. 19, 40–41 (1993).

    [227].    E.g., Derrick A. Bell, The Referendum: Democracy’s Barrier to Racial Equality, 54 Wash. L. Rev. 1, 1–7 (1978).

    [228].    See Washington v. Seattle Sch. Dist. No. 1, 458 U.S. 457, 470 (1982). But see Schuette v. BAMN, 572 U.S. 291, 322 (2014) (Scalia, J., concurring) (arguing that the “political-process doctrine” from Seattle Sch. Dist. No. 1 should be overruled).

It bears noting that the Court’s jurisprudence does not expressly state that the political process doctrine applies only in the context of direct democracy. Indeed, a recent case has sought to press the doctrine in the context of a majority-white state legislature in Alabama overriding the preferences of a majority-black city (Birmingham) with respect to minimum wage. Lewis v. Governor of Ala., 896 F.3d 1282, 1298 (11th Cir. 2018) (“The plaintiffs allege that the [state law] affects their ability to participate in the political process because it now occupies a field in which a majority-black [local] legislature previously enacted laws that they support.”), reh’g en banc granted, 914 F.3d 1291 (11th Cir. 2019).

It is also worth noting that the case most heavily relied on by the Court in Seattle School District No. 1 was Hunter v. Erickson, 393 U.S. 385 (1969), in which the Court held that an Akron, Ohio, city charter amendment, approved by voter initiative, that made it more difficult for the city council to enact local fair housing laws violated the Equal Protection Clause.  See 458 U.S. at 467–68 (discussing Erickson). As Erickson and Lewis demonstrate, therefore, the “political process doctrine” may apply at different levels: state voters overruling local law, state legislature overruling local law, and local voters constraining local lawmaking. See also Reitman v. Mulkey, 387 U.S. 369 (1967) (holding that voter initiative that amended constitution to prohibit laws against housing discrimination violated federal Equal Protection clause).

    [229].    See Daniel B. Rodriguez, State Constitutional Failure, 2011 U. Ill. L. Rev. 1243, 1268–73 (discussing the problems posed by voter-enacted tax and expenditure limitations in states).

    [230].    458 U.S. 457 (1982).

    [231].    Colo. Const. art. II, § 30b (“No Protected Status Based on Homosexual, Lesbian or Bisexual Orientation”).

    [232].    517 U.S. 620, 647 (1996).

    [233].    See generally David J. Barron, The Promise of Cooley’s City: Traces of Local Constitutionalism, 147 U. Pa. L. Rev. 487 (1999).

    [234].    Contra Lewis v. Governor of Ala., 896 F.3d 1282, 1298 (11th Cir. 2018) (majority-black Birmingham suing over majority-white Alabama state legislature’s preemption), reh’g en banc granted, 914 F.3d 1291 (11th Cir. 2019); cf. City of Richmond v. J.A. Croson Co., 488 U.S. 469, 495 (1989) (reviewing affirmative action plan of city with fifty percent black population and a majority-black city council).

    [235].    Campbell Gibson & Kay Jung, Historical Census Statistics on Population Totals by Race, 1790 to 1990, and by Hispanic Origin, 1970 to 1990, for Large Cities and Other Urban Places in the United States, tbl.48 (U.S. Census Bureau, Population Div., Working Paper No. 76, 2005) (indicating that Seattle’s population in 1970 was 87.4% white and 7.1% black).

    [236].    By “voluntary,” I mean an integration plan that was not imposed by a court in response to a desegregation lawsuit. See Paul Diller, Note, Integration Without Classification: Moving Toward Race-Neutrality in the Pursuit of Public Elementary and Secondary School Diversity, 99 Mich. L. Rev. 1999, 2000 n.4 (2001). While the Seattle plan at issue in Seattle School District No. 1 was not court-ordered, it was perhaps less than fully “voluntary” insofar as the school board adopted it in response to threats of segregation lawsuits by the local ACLU and NAACP chapters. Washington v. Seattle Sch. Dist. No. 1, 458 U.S. 457, 460 n.2 (1982) (“Several community organizations threatened legal action if the District did not initiate a more effective integration effort . . . .”).

    [237].    As of 1970, Washington was “overwhelmingly white,” with whites constituting 95.4% of the state’s population, and blacks 2.1% of total population. See John Caldbick, 1970 Census, (May 18, 2010), [https://perm].

    [238].    Seattle Sch. Dist. No. 1, 458 U.S. at 474; see also id. at 470 (“[T]he political majority may generally restructure the political process to place obstacles in the path of everyone seeking to secure the benefits of governmental action. But a different analysis is required when the State allocates governmental power nonneutrally, by explicitly using the racial nature of a decision to determine the decisionmaking process.”).

    [239].    Possibly. As noted above, the district had adopted the integration plan under threat of federal constitutional lawsuits by civil rights groups. See supra note 236.

    [240].    Seattle Sch. Dist. No. 1, 458 U.S. at 483.

    [241].    Romer v. Evans, 517 U.S. 620, 623–24 (1996) (citing each city’s ordinance).

    [242].    Id.

    [243].    Rodden, supra note 63, at 86–88 (describing the increasingly liberal social views of big cities as compared to rural areas).

    [244].    Romer, 517 U.S. at 634–35.

    [245].    Barron, supra note 233, at 594.

    [246].    See Linde, supra note 226, at 44.

    [247].    See Schuette v. BAMN, 572 U.S. 291, 318 (2014) (declining to overrule the political process doctrine, but not invoking it as requested by the plaintiffs).

    [248].    See Barron, supra note 233 and accompanying text.

    [249].    According to the Center for Immigration Studies, a nonprofit that is considered skeptical of heightened immigration, there are nine sanctuary states: California, Colorado, Connecticut, Illinois, Massachusetts, New Jersey, New Mexico, Oregon, and Vermont. Brian Griffith & Jessica M. Vaughn, Maps: Sanctuary Cities, Counties, and States, Ctr. for Immigr. Stud. (Apr. 16, 2019), (last updated Mar. 7, 2019) [].

    [250].    See Act of July 7, 1987, ch. 467, § 1, 1987 Or. Laws 914, (codified at Or. Rev. Stat. § 181A.820) (“No law enforcement agency of the State of Oregon or of any political subdivision of the state shall use agency moneys, equipment or personnel for the purpose of detecting or apprehending persons whose only violation of law is that they are persons of foreign citizenship present in the United States in violation of federal immigration laws.”).

    [251].    See Oregon Measure 105, Repeal Sanctuary State Law Initiative (2018), Ballotpedia [hereinafter Measure 105],,_Rep eal_Sanctuary_State_Law_Initiative_(2018) [].

    [252].    See id.

    [253].    For more on the details of “positive constitutionalism,” see Barron, supra note 233, at 493 n.15.

    [254].    Measure 105, supra note 251.

    [255].    According to 2018 Census estimates, Oregon was 75.3% non-Hispanic white and 13.3% Hispanic. Quick Facts: Oregon, U.S. Census Bureau (July 1, 2018), http://census. gov/quickfacts/OR [].

    [256].    See 2018 General Election Ballot Measures,, https://gov.oregon (displaying county-by-county vote for Measure 105) [].

    [257].    Compare id., with Portland State Univ., Census Data for Oregon, Percent Latino Population 2010, BY_CO.jpg [].

    [258].    See Sherman J. Clark, A Populist Critique of Direct Democracy, 112 Harv. L. Rev. 434, 469 (1998).

    [259].    See id. at 434.

    [260].    Ethan J. Leib, Can Direct Democracy Be Made Deliberative?, 54 Buff. L. Rev. 903, 911 (2006).

    [261].    See, e.g., Or. Rev. Stat. § 250.137 (establishing the “Citizens’ Initiative Review Commission” that can issue statements in favor or against ballot measures); see also John Gastil et al., Assessing the Electoral Impact of the 2010 Oregon Citizens’ Initiative Review, 46 Am. Pol. Res. 534, 554 (2018) (finding evidence suggesting that the citizens’ review process increased reflection and internal deliberation by voters); Ari J. Wubbold, Evaluating the Impact of Oregon’s Citizen Initiative Review (CIR) on Voter Decisions 4‒5, 75‒78, 80 (March 16, 2018) (unpublished M.A. thesis, Portland State University) (on file with the Portland State University Library), [].

    [262].    See Clark, supra note 258, at 434; Leib, supra note 260, at 903, 911; cf. Linde, supra note 226, at 34 (“Lawmaking by popular vote on an initiated proposal or course bypasses the committee study, hearings, amendments, and compromises of legislative deliberation . . . .”).

    [263].    David Magleby, Direct Legislation: Voting on Ballot Propositions in the United States 104–05 & tbl.6.1 (1984). For a more recent observation of a similar phenomenon, see Kerri Milita, Beyond Roll-Off: Individual-Level Abstention on Ballot Measure, 27 J. Elections Pub. Opinion & Parties 448, 463 (2017) (“[I]ndividuals most likely to abstain” on ballot measures have traits associated with “marginal voters—the young, the poor, and the non-white.”).

    [264].    Magleby, supra note 263, at 117.

    [265].    See, e.g., Benjamin I. Page et al., Democracy and the Policy Preferences of Wealthy Americans, 11 Persp. on Pol. 51, 51 (2013) (reporting that wealthy Americans are “much more conservative than the American public as a whole with respect to important policies concerning taxation, economic regulation, and especially social welfare programs,” and suggesting “that these distinctive policy preferences may help account for why certain public policies in the United States appear to deviate from what the majority of U.S. citizens wants the government to do”).

[266].    Property Tax Limitation, California Proposition 13 (1978) (codified at Cal. Const. art. XIII (A)).

    [267].    Id. (requiring two-thirds majority in both legislative houses for any tax increases).

    [268].    See Cal. Const., art. XIII A, §§ 1(a), 2(a)–(b) (restricting ad valorem property taxes to one percent of a property’s full cash value, and not allowing increases in valuation more than two percent, except upon transfer of ownership); see also Nordlinger v. Hahn, 505 U.S. 1, 28–30 (1992) (Stevens, J., dissenting) (discussing the inequities in property taxation imposed by Prop 13).

    [269].    See generally Erin Adele Scharff, Powerful Cities? Limits on Municipal Taxing Authority and What To Do About Them, 91 N.Y.U. L. Rev. 292 (2016).

    [270].    State & Local Tax Revolt: New Directions for the ‘80s, at 80 (Dean Tipps & Lee Webb eds., 1980).

    [271].    Id. at 78–79.

    [272].    Nat’l Conference of State Legislatures, Preparation of Fiscal Analysis (Apr. 2002), px (discussing the important educational role that “fiscal impact statements” play in the thirteen states that have some version of the requirement for ballot measures) [https: //].

    [273].    For more on this measure, see supra notes 249–52 and accompanying text.

    [274].    Or. Sec’y of State, Oregon Measure No. 105, in Voter’s Pamphlet: Or. Gen. Election, Nov. 6, 2018 at 111, 112–13, [].

    [275].    Multnomah Cty. Bd. of Comm’rs, Res. 2016-132 (Or. 2016), file/58697/download []; see QuickFacts: Multnomah County, Oregon, U.S. Census Bureau (July 1, 2018), mahcountyoregon []; QuickFacts: Oregon, U.S. Census Bureau (July 1, 2018), [ PD].

    [276].    Peter Wong, Clackamas County Reaffirms Support of Diversity, Hands-Off Immigration Actions, Portland Trib. (June 8, 2017), 671-242830-clackamas-county-reaffirms-support-of-diversity-hands-off-immigration-act ions [].

    [277].    See, e.g., Susan Kass, Opinion, Passing Measure 105 Would Send a Chilling Message to Children, (Oct. 7, 2018), 18/10/opinion_passing_measure_105_wo.html [].

    [278].    See, e.g., Town Hall on Measure 105: Should Oregon Repeal Its Status as a Sanctuary State?, KATU News (Oct. 15, 2018), ure-105-should-oregon-repeal-its-status-as-a-sanctuary-state [ PZ]; Richard F. LaMountain & John Ratliff, Opinion, Two Takes on Ballot Measure 105, Oregonian, .html (last updated Jan. 29, 2019) []; Henry Houston, Eugene City Council Announces Opposition to Measure 105, Eugene Wkly. (Oct. 29, 2018) http:// -105/ [].

    [279].    Cf. Davidson, supra note 6, at 993 n.144 (noting the role, in the context of constitutional rights, for “higher-level governments to set a floor, above which local governments can craft policy”).

    [280].    See generally Elisabeth R. Gerber, The Populist Paradox: Interest Groups Influence and the Promise of Direct Democracy 4–5 (1999) (recounting “interest groups,” criticism of direct democracy); see also David S. Broder, Democracy Derailed: Initiative Campaigns and the Power of Money (2000).

    [281].    To be sure, there may be some “traditional” lobbying of elected officials to get them to endorse a particular side in an initiative fight. See Arthur Lupia & John G. Matsusaka, Direct Democracy: New Approaches to Old Questions, 7 Ann. Rev. Pol. Sci. 463, 471–72 (2004) (“Convincing voters that an initiative represents an improvement over the known status quo . . . requires more than money. . . . [i]t also requires the endorsements of well-known public figures and evidence of broad grass-roots support.” (citations omitted)).

    [282].    Id. at 471 (“A fact beyond dispute is that qualifying a measure for the ballot can be expensive.”); e.g., Rob Davis, Ballot Initiative To Tighten Oregon Forestry Law Gets Rejected. Advocates Blame Timber Money., OregonLive (Oct. 4, 2019), environment/2019/10/ballot-initiative-to-tighten-oregon-forestry-laws-gets-rejected-advoca tes-blame-timber-money.html (discussing example of the Oregon Secretary of State’s office refusing to certify ballot measures because it believed they violated the state constitution’s “single-subject” requirement) [].

    [283].    In 2018 in California, for instance, the opposition to an initiative that would have limited profits of dialysis centers spent $111 million to defeat it. See Laurel Rosenhall, Record Spending as Huge Money Flows into Industry Fights on the California Ballot, KQED News (Nov. 4, 2018), -flows-into-industry-fights-on-the-california-ballot [].

    [284].    David F. Damore & Stephen P. Nicholson, Mobilizing Interests: Group Participation and Competition in Direct Democracy Elections, 36 Pol. Behav. 535, 549 (2014).

    [285].    Leib, supra note 260, at 907.

    [286].    Gerber, supra note 280, at 119 (concluding that “economic groups find it very difficult to pass new initiatives”); id. at 138–39 (“Economic interest groups devote a much greater share of their financial resources to campaigns to oppose initiatives than to campaigns to support initiatives, and the measures they oppose fail at a high rate.”).

    [287].    Act of Aug. 3, 1995, ch. 331, 1995 Cal. Stat. 1819, 1820 (codified at Cal. Civ. Code §§ 1954.50–.535).

    [288].    California Proposition 10, Local Rent Control Initiative (2018), Ballotpedia,,_Local_Rent_Control_Initiative_(2018) [].

    [289].    Id.

    [290].    Id.

    [291].    Indeed, the Supreme Court has struck down attempts to limit individual and corporate contributions to ballot initiative campaigns. See Citizens Against Rent Control v. City of Berkeley, 454 U.S. 290, 300 (1981) (striking down California law that set limits on contributions to initiative campaigns); First Nat’l Bank of Bos. v. Bellotti, 435 U.S. 765, 776 (1978) (holding that corporations have a First Amendment right to contribute to ballot initiative campaigns). Contribution limits to individual candidate campaign accounts, by contrast, remain on the books in many states, even if the Supreme Court has overall grown more hostile to campaign finance regulation. See State Limits on Contributions to Candidates, 2017–18 Election Cycle, Nat’l Conf. St. Legislature, (June 27, 2017), http://www. 16465.pdf (showing that thirty-nine of fifty states limit individual contributions to candidates’ campaigns, and forty-five limit or prohibit corporate contributions to candidates’ campaigns) [].

    [292].    Kenneth T. Andrews & Bob Edwards, Advocacy Organizations in the U.S. Political Process, 30 Ann. Rev. Soc. 479, 481 (2004) (discussing definitions of “interest groups”).

    [293].    Id.

    [294].    Id.

    [295].    Kenneth Andrews and Bob Edwards define “advocacy organizations” as those that “make public interest claims either promoting or resisting social change that, if implemented, would conflict with the social, cultural, political, or economic interests or values of other constituencies and groups.” Id.; see also Matt Grossmann, The Not-So-Special Interests 24 (2012) (discussing “advocacy organizations”).

    [296].    See Alexandra Sifferlin, Soda Taxes Prompt High Fives from Health Advocates, (Nov. 11, 2016), (reviewing the soda taxes passed by four cities and one county in November 2016) [].

    [297].    See Seattle, Wash., Ordinance No. 125324 (codified at scattered sections of Seattle Mun. Code 5.30, 5.53, 5.55).

    [298].    See Notice of Multnomah County Initiative Ballot Title & Explanatory Statement, MultCoInit-04, Multnomah County (2017), ah-co-initiative-ballot-title-explanatory-statement-multcoinit-04 []; see also Nigel Jaquiss, Backers of Multnomah County Soda Tax Delay Again, Will Not Pursue November Ballot Spot, (Feb. 23, 2018), https://www. [].

    [299].    See Initiative and Referendum States, supra note 219.

    [300].    See Initiative Measure No. 1634 (2018), Wash. Secretary St., https://www.sos. [].

    [301].    See Washington Initiative 1634, Prohibit Local Taxes on Groceries, Ballotpedia,,_Prohibit_Local_Taxes_on_Groceries_ Measure_(2018) []; Yes! to Affordable Groceries, Pub. Disclosure Commission, filer_id=YESTA%20%20015&election_year=2018 [].

    [302].    See Yes! to Affordable Groceries, supra note 301.

    [303].    See Washington Initiative 1634, Prohibit Local Taxes on Groceries, supra note 301.

    [304].    See Initiative, Referendum, and Referral Search, Initiative Number 37, Or. Secretary St., 20180037..LSCYY.Y [].

    [305].    See November 6, 2018, General Election Abstract of Votes, Or. Secretary St., [].

    [306].    Oregon Measure 103, Ban Tax on Groceries Initiative (2018), Ballotpedia, https: //,_Ban_Tax_on_Groceries_Initiative_(2018), [https://].

    [307].    See supra note 285 and accompanying text (discussing how it is easier to buy a “no” than a “yes”).

    [308].    See Online Voters’ Pamphlet, 2018 General Election, Or. Secretary St., http://oreg [https://perma. cc/22GW-VMUP].

    [309].    See Hillary Borrud, Labor Groups Link Kate Brown’s Re-election, Initiatives’ Defeat, (Nov. 4, 2018), ps_link_kate_browns.html (discussing the broad, union-funded coalition that worked to defeat Measure 103, among its many causes in the 2018 election cycle) [].

    [310].    Editorial Endorsement: Vote “No” on Measure 103’s Grocery-Tax Ban, (Sept. 30, 2018), endorsement_vote_no.html (objecting that Measure 103 “would bar local communities from enacting a targeted food or beverage tax of their own, even if their community members enthusiastically support one”) [].

    [311].    Sign on file with author.

    [312].    Disappointingly, for those who advocate local fiscal control, despite the successful effort to fend off Measure 103 in the November 2018 election, the legislature then imposed the same sort of preemption during its 2019 legislative session. Act effective Sept. 29, 2019, ch. 579, § 67, 2019 Or. Laws (“A city, county, district or other political subdivision or municipal corporation of this state may not impose, by ordinance or other law, a tax upon commercial activity or upon receipts from grocery sales.”).

[313].    See Paid Sick Time Laws, Overview of Paid Sick Time Laws in the United States: Oregon, Better Balance, (last updated May 20, 2019) [].

    [314].    See Diller, supra note 60, at 1129 (discussing this dynamic).

[315].    See id. In apparent deference to Portland’s ordinance, the legislature allowed the city’s more generous threshold for employees (six, as opposed to the state law’s ten) to apply to businesses with a presence in Portland. Id.

[316].    The converse might occur when cities or their allies seek to repeal pre-existing blanket preemption, and, therefore, must make significant sacrifices to their own autonomy going forward just to regain some of it. Oregon’s law allowing for inclusionary zoning, passed in 2016, is a good example in this regard. See Jeff Mapes, Oregon Legislators Reach Deal on Affordable Housing Legislation, OPB (Feb. 23, 2016, 3:38 PM), article/oregon-legislators-reach-deal-on-affordable-housing-legislation (last updated Feb. 24, 2016, 2:30 PM) (noting that “[t]he ban preventing local governments from using inclusionary zoning was approved by the 1999 Legislature at the behest of the Oregon Home Builders Association,” but that “the home builders and other development groups agreed to [lift the ban] after they won concessions limiting how local governments can use inclusionary zoning”) [].

[317].    Indeed, some of these counties successfully challenged the paid-sick-leave law as applied to them as an unconstitutional unfunded mandate. Linn Cty. v. Brown, 443 P.3d 700, 702 (Or. Ct. App. 2019).

[318].    Briffault et al., supra note 1, at 1997; Scharff, supra note 1, at 1472.

[319].    See, e.g., United States v. Carolene Prods. Co., 304 U.S. 144, 152 n.4 (1938) (suggesting “more searching judicial inquiry” when legislation is motivated by “prejudice against discrete and insular minorities”).

[320].    City of El Cenizo v. Texas, 264 F. Supp. 3d 744, 809, 812 (W.D. Tex. 2017) (enjoining on First Amendment grounds the section of Texas’s anti-sanctuary-city law providing for removal from office of local officials who “endorse” sanctuary policies), aff’d in part, vacated in part, 890 F.3d 164, 192 (5th Cir. 2018); Briffault et al., supra note 1, at 2010–11.

[321].    In the November 2018 elections alone, five states’ voters—Colorado, Michigan, Missouri, Ohio, and Utah—passed ballot measures to establish more neutral methods for legislative districting. Robert Barnes, Supreme Court Again Considers Partisan Gerrymandering, but Voters Are Not Waiting, Wash. Post (Mar. 24, 2019), https://www.washington [].

[322].    Rucho v. Common Cause, 139 S. Ct. 2484, 2495–96, 2507 (2019). On recently successful state constitutional attacks on political gerrymandering, see supra note 218.

[323].    See supra note 291.

    [324].    E.g., Ill. Const. art. XIV, § 3 (limiting voter initiatives to “structural and procedural subjects”).

[325].    See Paul A. Diller, Reorienting Home Rule: Part 2—Remedying the Urban Disadvantage Through Federalism and Localism, 77 La. L. Rev. 1045, 1077–95 (2017) (arguing that home rule can serve as a remedy for the disadvantages urban areas suffer in the state political process).