What Can Latin America Tell Us about Subnational Democratic Erosion in the United States?

Abstract

This article uses the literature on subnational undemocratic regimes (SURs) and regime juxtaposition in Latin America to gain analytical leverage on the recent process of subnational democratic erosion in the United States. Based on a review of five key dimensions of federalism, we argue that the institutional landscape for the emergence and continuity of SURs is, comparatively speaking, more favorable in the US than in any of Latin America’s three federations (Argentina, Brazil, and Mexico). In addition to showing how federal structures are more detrimental for subnational democracy in the US, we assess how the two main theoretical approaches that have been developed to understand SURs in Latin America and elsewhere can be applied to the US.

Recent years have witnessed serious instances of democratic erosion at the subnational (i.e., state) level in the United States. Scholars of the US have documented a number of worrisome developments. These include extreme gerrymandering to stack the deck in favor of one party in state legislatures, efforts by those legislatures to then disempower independent judiciaries and governors from the opposition party, and new rules that impose significant obstacles on voters seeking to exercise their democratic rights (Grumbach 2022; Levitsky and Ziblatt 2023; Lieberman, Mettler, and Roberts 2021; Mettler and Lieberman 2020; Stephanopoulos and Warshaw 2020). This erosion in the US is taking place against the backdrop of the subnational undemocratic regimes (SURs) that operated during the long decades of Jim Crow, and which came to an end only 50 years ago (Mickey 2015). While the phenomenon of subnational democratic backsliding is now cause for concern throughout the country and not just in the 11 states of the former Confederacy (Grumbach, Hacker, and Pierson 2021), there are also good reasons to anticipate that contemporary democratic erosion will not result in the reemergence of fully fledged SURs (Mickey 2022). A better understanding of what has changed across these time periods in the US, and what has not, has become a research imperative and the focus of much important recent work on the US, one that makes possible a new cross-temporal research agenda on subnational political regimes (Grumbach and Michener 2022; Mickey 2022; Mickey, Levitsky, and Way 2017; Rocco 2021).

In this article, however, we engage in a comparison of a different sort. Rather than comparing the US across time periods, we seek to broaden our understanding of this phenomenon by adopting a cross-regional perspective. Here we are interested in asking what the SUR research agenda in Latin America can tell us about what is happening in the US. Our comparative focus on Latin America is driven by the vast and compelling work on this subject. In fact, this world region has produced by far the largest body of scholarship on the topic, ¹ starting with the influential work of Guillermo O’Donnell (1993), who more than 30 years ago identified the existence of so-called brown areas within new Latin American democracies. O’Donnell defined these areas as regions characterized by a lack of the rule of law and the absence of democracy. Drawing on this seminal notion, a robust literature on subnational democracy (and the lack thereof) emerged in the region to refine the conceptualization of SURs, measure and compare them, and explain their endurance despite national democratization. ² This prolific body of literature on regime juxtaposition has centered for the most part on accounting for the causes that led to instances of subnational democratic erosion or SUR reproduction in Latin American federations, most of which had transitioned to liberal democracy at the national level in the 1980s, 1990s, and early 2000s. Back then, the bulk of these SURs experienced what we observe today in some US states: a sustained erosion of democratic principles and institutions. ³ Despite similarities in the process of subnational democratic erosion, these countries have seldom been compared (cf. Gibson 2013).

For the purpose of this article, we focus on Latin America’s federal countries, ⁴ and ask what we see when we compare SURs in Argentina, Brazil, and Mexico with the present-day US. Although the current literature on the US is centered on subnational democratic erosion while the literature on Latin America has focused on erosion, SUR continuity, and subnational democratization, much can be learned by comparing how “varieties of federalism” either empower or constrain would-be subnational autocrats and their parties. Analyzing this phenomenon across the Americas is facilitated by a number of important institutional similarities, especially the combination of presidentialism and federalism. Despite a lack of theoretical consensus about the survival or transformation of SURs, a review of what has been learned about this phenomenon in Latin America over recent decades points to a striking pattern: on several key dimensions, the institutional landscape for the emergence and continuity of SURs is, comparatively speaking, more favorable in the US than it is in Latin America’s three federations. Turning on its head the common wisdom that democracy and democratization face greater challenges south of the border than they do north of it, we draw attention to five key institutional dimensions.

First, while subnational democratic erosion has occurred through attempts to limit partisan contestation in the US and in Latin America’s federal countries alike, only in the former case do we also see attempts to formally limit participation through explicit projects of voter suppression, which are largely missing from SURs in Argentina, Brazil, and Mexico. Second, subnational autocrats in the latter three federations have sought to build SURs from all ideological points in the political spectrum, whereas in the US the subnational autocratic impulse has largely emanated from a single party (the Democratic Party under Jim Crow and the Republican Party in the contemporary period). Third, the US lacks the institutional mechanism—that of federal intervention—that has sometimes enabled democratic presidents to remove subnational autocrats in Argentina and Mexico. Fourth, whereas SURs in Latin America have themselves often been transformed in a democratic direction due to municipal pressures from below, in the US the norm of state preemption vis-à-vis municipalities likely weakens this possibility as a “bottom-up” mechanism of democratic resistance in the face of state-level backsliding. Finally, no federation in Latin America (indeed no country in the world) gives subnational jurisdictions the kind of control over federal elections that is enjoyed by the 50 states in the US. As a result, SURs in the US can do more to effect national-level backsliding than would be possible anywhere in Latin America’s federations.

Before further unpacking each of these five institutional dimensions, we briefly review the Latin American literature that has conceptualized SURs. In the last section, we succinctly discuss how the two most common theoretical approaches that have been developed for Latin America (and the Global South) can be applied to the study of SURs in the US. The article concludes with some reflections about future research on subnational and national democratic erosion.

What Are SURs?

Conceptualizing Latin American SURs has generated a great deal of scholarly debate, including about whether these regimes are actually possible in countries that are regarded as national democracies (Behrend and Whitehead 2016), and whether they should be referred to as authoritarian (Gibson 2013), hybrid (Gervasoni 2018), undemocratic (Giraudy 2010; 2015), or illiberal (Behrend and Whitehead 2016). The general agreement is that these subnational regimes do not comply with the minimal, procedural definition of democracy. When it comes to the electoral component of democracy, SURs fail to guarantee key electoral processes, such as regular, free, and clean elections, and violate core political and civil rights for voters and opposition parties. Contrarily, disagreement on SUR conceptualization typically revolves around whether additional dimensions of democracy (i.e., the liberal and majoritarian components) need to be present for a regime to classify as a SUR. Some of these include weak checks and balances, the existence of family networks, and/or the presence of patrimonial state administrations. In sum, most scholars of SURs in Latin America agree that at a bare minimum, SURs severely restrict the electoral dimension of democracy—a definitional stance that would surely apply to the US as well. ⁵

Comparing the US with Latin American Federations across Five Institutional Dimensions

Voter Participation

If contestation and participation are frequently conceptualized as the two main dimensions of electoral democracy (Dahl 1971), one of the most intriguing observations about SURs in Latin America is that they have only affected the first dimension. Subnational autocratic incumbents in Argentina, Brazil, and Mexico have engaged in efforts to suppress contestation through their control over electoral commissions, their ability to gerrymander districts to limit representation by opposition parties, and their imposition of limits on freedom of organization (Calvo and Micozzi 2005; Gibson 2013; Giraudy 2015). Suppressing the opposition in some cases has been enough to enable governors to perpetuate themselves (or their kin) in power in the form of long-term rule by provincial family clans (Behrend 2016). What they have not sought to suppress, however, is voter participation through legal changes that would restrict who can vote in provincial elections. Voters may effectively experience limited choices given the success of measures that weaken contestation, but they do not generally have cause to fear being legally unable to vote. The lack of systematic voter suppression in Latin America appears to have a number of causes, including the strong regionwide norm of mandatory voting and extensive practices of clientelism that may obviate the need to suppress votes that can be purchased cheaply.

By contrast, subnational democratic erosion has occurred in both the contestation and participation dimensions in the US, where disenfranchisement of voters is today quite prevalent. Several studies have shown that voter suppression and changes designed to slow or stop the expansion of the electorate are common characteristics of American SURs (Grumbach and Hill 2022; Mickey 2022). Beginning in the early 2000s, attempts to shrink the voting population and make it harder to vote have increased significantly, with the Supreme Court’s 2013 decision vis-à-vis the 1965 Voting Rights Act constituting the watershed event. Almost half of all US states have introduced restrictions on eligibility and turnout in the years since Shelby County v. Holder. According to Mickey (2022, 122), “restrictions on voter eligibility include photo identification requirements and purges of registration rolls,” and “restrictions on turnout include reducing early voting options, drastically reducing the number of polling places, and reducing extensions of polling hours.” As Hill (2020) argues, these attempts are often targeted to limit turnout among particular kinds of voters, including younger voters, voters of color, and immigrants. We note here the debate among Americanists over both the extent of voter suppression and the impact that it might have on democracy. Some question whether voter ID policies have actually reduced turnout and suggest that merely blocking reforms that would expand participation is not technically backsliding, while others argue that voter suppression may not require observable reductions in overall turnout. ⁶

The importance of systematic attempts to suppress participation in the US and the absence of such attempts in Latin America underscore the more extreme nature of democratic erosion in the former case. It also means that the process of subnational democratic backsliding may be harder to check because it takes more diverse forms, including perfectly legal measures. Outside allies in the struggle against erosion, who have proven to be so critical in the promotion of more democratic subnational outcomes in Latin America (Gibson 2005), likely have an easier time finding ways to help local opposition groups to contest elections than undoing legal changes that directly restrict the electorate.

Ideological Orientation

In Latin America, SURs have emerged at all points on the ideological spectrum, a pattern that replicates dynamics at the national level where some authoritarian regimes have expressed a hard-right political orientation (e.g., in Argentina and Chile) while others have espoused much more progressive ideals (e.g., in Bolivia, Peru, and Ecuador). Most of the SURs that emerged and consolidated in Argentina and Brazil in the 1980s and 1990s, and in Mexico after 2000, cannot be identified with one single ideology. SURs in Argentina have been left leaning, as in the case of the hegemonic Neuquén People’s Movement (Movimiento Popular Neuquino, MPN) party in the province of Neuquén, for example. Likewise, just within the subset of Peronist SURs, both conservative and right-wing cases exist, like in Santiago del Estero and Formosa provinces, and center-left examples abound, as in the case of San Luis province (Gervasoni 2018; Gibson 2013; Giraudy 2015). A similar ideological diversity can be seen in Mexico in the early 2000s, where SURs existed under the rule of the center-left Institutional Revolutionary Party (Partido Revolucionario Institucional, PRI) in the states of Oaxaca and Puebla, for instance; the right-wing National Action Party (Partido Acción Nacional, PAN) in Jalisco state; and in left-leaning Party of the Democratic Revolution (Partido de la Revolución Democrática, PRD) districts, such as in Tabasco state (Giraudy 2015). The same is true in Brazil, where SURs in the 1980s and 1990s prevailed under the leadership of the right-wing Democrats (DEM), formerly the Liberal Front Party (Partido da Frente Liberal, PFL), as was the case in the state of Bahia; or under the rule of the center-right Brazilian Democratic Movement Party (Partido do Movimento Democrático Brasileiro, PMDB), as happened in the state of Rio Grande do Norte (Borges 2016; Souza 2016). Rulers from all of these parties have worked to suppress party contestation and erode checks and balances, and political parties of all colors have engaged in some sort of political violence to deter opponents from taking state power.

In the US, by contrast, subnational democratic erosion has taken place under the auspices of a single party, though the identity of the responsible party has changed over time from Democratic to Republican. Before the 1960s, SURs existed in the US South under the dominance of the Democratic Party through a party-specific toolkit that included the white primary (Mickey 2015). More recently, the Republican Party has been the main culprit vis-à-vis democratic erosion (Grumbach 2022; Grumbach and Hill 2022; Mickey 2022). In recent years, Republicans have engaged in aggressive attempts to undermine checks and balances in states like Georgia, North Carolina, and Wisconsin. According to Grumbach (2022) the magnitude of democratic contraction under Republican control is remarkably large, and it has taken place through gerrymandering and changes in election policies after Republicans made gains in state legislative and gubernatorial elections in 2010. While Democrats have sought to make voting easier and to broaden the electorate, Republicans have been actively implementing reforms to suppress voting—especially by presumed Democratic voters (Brennan Center for Justice 2021).

The greater ideological heterogeneity of SURs in Latin America and the reality that they exist under a broader set of political conditions could mean that attempts to undermine them may theoretically face a more varied set of hurdles. Overall, however, the dynamic we see in the US is likely to be far more troubling. Where subnational democratic erosion can largely be attributed to just one partisan actor, the combined impact of these efforts at the national level may be more significant and destabilizing, especially during periods of heightened partisan polarization. As Grumbach (2022) argues, US states have emerged as “laboratories against democracy” due to national-level gridlock that has encouraged the Republican Party to shift its focus to the states, where state-level party organizations are drawing from a common playbook.

Here it is important to note the contrast between robust SURs under Democratic auspices during Jim Crow, when the national Democratic Party was ideologically diverse internally, with a contemporary period marked by a more ideologically homogeneous Republican Party in which we do not (yet) see fully fledged SURs. Similar and concerted attempts by one increasingly extreme ideological actor (the Republican Party) to suppress the vote in otherwise quite different states (e.g., Wisconsin, North Carolina, Texas, and Ohio) likely increase the national import of this subnational trend relative to what we are seeing in Latin America, where ideologically disparate agents of backsliding do not coordinate their actions or seek to scale up their attempts at erosion. In Mexico, to name one example, highly disparate multiparty coalitions across the 32 states are heterogeneous in ways that inhibit their uniform impact on the national level.

Federal Intervention

One important potential check on SURs in federal countries is the possibility that the central government can suspend a provincial or state government and govern that subnational unit directly, which is referred to as “federal intervention” in Latin America. Democratic presidents in Argentina and Mexico can use, and have used, this authority to remove autocratic governors from office. The relevant constitutional language justifies these interventions if they are deemed necessary to preserve the “republican form of government.” Nothing guarantees that this rule, if it exists, will be used for democracy-enhancing purposes; indeed, there may be important cases in which the persistence of SURs redounds to the political benefit of democratically elected presidents (Giraudy 2015). Furthermore, as Behrend (2016) has found in the Argentine case, the removal of an individual authoritarian governor has not necessarily guaranteed that subnational democratization has followed instead of an equally authoritarian successor. Nevertheless, the rule can be and has been used to remove from office those who have perpetrated subnational democratic erosion. While federal intervention has been used heavily in the Argentine and the Mexican cases (Behrend 2016; Gibson 2013; Giraudy 2015) it has not been used in Brazil since redemocratization despite the fact that it is indeed a constitutional prerogative of the federal government (Souza 2016, 201). Recently, the Supreme Federal Court in Brazil temporarily removed the governor of Brasília for failing to stop the violent attack on January 8, 2023 against the three branches of the federal government.

Reflecting its “coming together” origins as a federation in which the US states retained important forms of sovereignty under the 1789 Constitution, the federal institutional landscape is quite different in the US. While federal intervention was arguably the single most important factor in the ending of Jim Crow and in the timing of subnational democratization in the South (Gibson and King 2016; Mickey 2015), and while the US Constitution “mandates that the central government guarantee a ‘republican form of government’ in every state vis-à-vis threat of invasion” (Gibson 2013, 76), it does not give the federal government the right to suspend state governments as in Latin America. Action at the federal level to democratize the South took a number of forms, including executive-branch decisions to federalize the national guard in Arkansas, Supreme Court decisions that instructed lower courts to rule consistently with Brown v. Board of Education in lawsuits that were filed against school districts refusing to desegregate, and the congressional passage of the 1964 Civil Rights Act and 1965 Voting Rights Act. None of these mechanisms, however, could be used to remove individual subnational autocrats from office, which is an important prerogative for the federal government, even where it mostly works as a threat rather than as a rule that is frequently deployed. If Republican efforts to erode subnational democracy were to become even bolder over time, democratically elected politicians in charge of the federal government would be denied the use of an important tool that their counterparts in Latin America have been able to wield to defend subnational democracy.

Municipal Autonomy

In Latin America, municipal governments have played critical roles as spaces of resistance vis-à-vis autocratic governors, a dynamic that Ed Gibson (2005, 131) refers to as the “authoritarian province, plural cities” phenomenon. In his formulation, democratic actors in politically competitive cities can initiate attempts to draw the federal government into their disputes with nondemocratic governors, who instead implement strategies of “boundary control.” In the Mexican case, for example, Gibson (2013, 132) argues that “urban hubs” served as the launching pads for anti-SUR struggles, and notes that “in 15 of 18 party-led transitions between 1997 and 2010, opposition parties won the governorship [from the PRI] after having held the mayoralty of a major urban hub.” According to Celina Souza (2016, 222), the demise of authoritarian rule in Brazilian states like Bahia (dominated for 27 years by Governor Antônio Carlos Magalhães) was facilitated by a key change in the 1988 Constitution, which gave municipalities a separate status from the states where they are located. This status, combined with decentralization, enabled municipalities to partner with the national government in offering services that made voters less dependent on the governor’s clientelistic appeals (Montero 2012). We see a similar dynamic in Argentina, despite the fact that, like the US and unlike Brazil and Mexico, the rights and responsibilities of municipal governments are at the discretion of the provinces and established in provincial constitutions (rather than the federal constitution). For instance, in San Luis province, autocratic governors (and brothers) Adolfo and Alberto Rodríguez Saá were constrained by a provincial law that automatically shares revenues with municipalities, including those governed by opposition mayors (Giraudy 2015, 113).

In the US, in contrast, not only do undemocratic governors have less cause to fear removal from above than their counterparts in Latin America (due to the absence of an institutionalized mechanism of federal intervention in the US), but they also have less reason to fear the effects of democratizing pressures from below—that is, from the local (municipal) level. Like many countries in the world, the US contains extensive examples of juxtaposition in which rural and conservative interests easily dominate state-level institutions in states with cities whose municipal governments seek to pursue a more inclusive agenda (Mettler and Brown 2022; Rodden 2019). The US Constitution gives cities and counties very little cover should governments at this local level adopt policies or behaviors that challenge SURs at the state level; instead, states are able to engage in “preemption” to foreclose these challenges. Examples abound in recent years of this state–municipality conflict, and of the stark reality that municipal governments are mere “creatures of the states.” For example, Republican lawmakers in Georgia in 2021 followed up their passage of a new and more restrictive election law with a request to review the handling of elections by Democratic stronghold Fulton County (home to 11% of the state’s electorate). Also in 2021, Texas governor Greg Abbott signed a bill that limited the ability of counties to expand voting options, which Democratic Harris County and other counties had discussed in response to earlier state-level restrictions. Two years later, Texas Republicans passed House Bill 2127, which constrained powers of self-governance in cities whose liberal policy orientation (from labor to agriculture) conflict with the conservative statewide agenda. Also in 2023, the National Association for the Advancement of Colored People sued the state of Mississippi to stop the passage of two new laws that would give the state more control over the capital city of Jackson, including rules that would prevent residents in that city from demonstrating in and around buildings considered property of the state. In Latin American federalism, not only is this kind of preemption far less salient as an institutional practice, but it is also not generally characterized by the racial dynamics that seem to fuel preemptive moves against Black-majority urban centers in the US.

Subnational Role in Federal Elections

The final important comparative dimension is perhaps the most worrisome in terms of the much greater national-level impact that subnational democratic erosion can have in the US as compared to anything that we see in Latin America. In Latin America, national elections are under the control of national authorities. Antidemocratic actors in Argentina, Brazil, and Mexico can do real damage to democracy subnationally, but in no case have they ever been able to regulate the federal electorate. In the Mexican case, the transition to democracy in the 1990s significantly strengthened the ability of the Federal Electoral Institute (Instituto Federal Electoral) to monitor elections to federal offices in the states (Gibson 2013), and concerns more recently over undemocratic actions by Mexican governors led to reforms in 2012 (under the so-called Pact for Mexico) that also nationalized control over state-level elections (Baez 2019). In Argentina, partisan representatives can monitor elections for federal office in provincial districts, but provincial governments themselves do not play a role in national-level elections (Gibson 2013, 124). Also, with the abolition of the Argentine electoral college in that country’s constitutional reform in 1994, no other Latin American federation gives their states or provinces this potentially important role in electing the president. National control over national elections in Latin America has facilitated some of the most technologically sophisticated voting systems anywhere in the world, including Brazil’s pioneering effort to register all voters through biometric identification, which aims to achieve 100% coverage by the 2026 elections.

In contrast, one key difference that distinguishes the US from all of the Latin American cases is the rule that gives the 50 states significant prerogatives vis-à-vis not just state and local but also federal elections. According to the highly decentralized approach to federal elections in the US, the constituent units of the federation control the (re)districting of all seats in the lower chamber as well as rules governing the eligibility to vote in federal elections, and state legislatures play a potentially critical role in the certification of presidential election results. In the Jim Crow past, administrative control over federal elections extended tremendous leverage to SURs in the US (and possibly will do again in the future). During the many decades of the Solid South, SURs in the former Confederacy disenfranchised millions of Black voters in ways that severely reduced the effective constituency for progressive policy change at the federal level, especially since SUR governors belonged to the very (Democratic) party from which most such policy proposals could be expected to emanate. In addition to the possibility of manipulating who can vote in federal elections, the US further empowers the states by using an electoral college to select the president, with state legislatures able to determine how these electors are selected. In the years to come, manipulation of the electoral college by antidemocratic forces at the subnational level will pose perhaps the greatest threat to national-level democracy in the US.

Theorizing Subnational Undemocratic Regimes in the Global South and in the US

If, as argued above, a comparative assessment in the Americas draws our attention to the especially permissive institutional environment in the US for SUR consolidation, what can it tell us about how this phenomenon should be theorized? Here we take stock of the vast body of works produced by scholars of SURs in the Global South to identify the main analytical approaches used to study the origins and continuity of these nondemocratic regimes.

Broadly speaking, two different analytic and theoretical approaches exist as of today. The first approach, referred to as the unilevel approach (Giraudy, Moncada, and Snyder 2019), focuses on subnational variables as the main drivers of SUR origins, reproduction, and consolidation. That is, only variables at the same scale (i.e., level of government) as the outcome of interest are understood to have a causal effect (while allowing that variables at higher or lower scales may provide contextual information). In this approach, second-tier level-of-government variables, such as the economic autonomy of inhabitants (McMann 2006), levels of poverty and the spatial location of clientelistic machines in each state or province (Montero 2012), a SUR’s geographic location and whether it is close to a neighboring country (Lankina and Getachew 2006), the level of local citizens’ human capital (Lankina and Getachew 2012), and the size of SURs’ electoral districts (Gerring, Teorell, and Zarecki 2013) are the main predictors of SUR origins and continuity. One example of this unilevel approach in the US is Robert Mickey’s influential Paths out of Dixie (2015), which explains distinct patterns of subnational democratization as the result of state-level differences in elite cohesion and party-state institutions. In sum, according to this first approach, a variety of subnational factors specific to each subnational unit accounts for the origins, persistence, and consolidation of SURs.

A second approach to the study of SURs instead views subnational political outcomes as a byproduct of the political dynamics that play out at the intersection of national- and subnational-level arenas. Works within this second approach, the so-called multilevel approach (Giraudy, Moncada, and Snyder 2019), build on the premise that, in large-scale systems of territorial governance, political institutions are entangled across space—and precisely for that reason, political action and political outcomes, such as the origin and maintenance of SURs, are not limited to a single arena (Rokkan 1970; Rokkan and Urwin 1982; 1983; Tarrow 1978). On the contrary, subnational political outcomes are the product of regular interventions on the part of national governments and national institutions, such as political parties, territorial regimes, or fiscal arrangements. Proponents of a multilevel approach in the Global South argue that SURs are shaped by presidents’ strategic interactions with these regimes (Gibson 2005; 2013), the ability of presidents to muster fiscal or partisan control over SURs (Giraudy 2013; 2015), national-level policies such as neoliberalism (Snyder 2001) or decentralization reforms (Montero and Samuels 2004), and national- and municipal-level fiscal institutions (Díaz-Rioseco 2016; Gervasoni 2010; 2018). One pathbreaking example of this multilevel approach in the US is Jacob Grumbach’s Laboratories against Democracy (2022), which emphasizes the importance of the racial, geographic, and economic incentives of groups in national party coalitions. Multilevel approaches are often more realistic, but can be harder to pull off given the presence of both vertical and horizontal feedback loops.

Each of these theoretical approaches, which were largely developed to understand subnational democratic erosion in the Global South, can indeed help to illuminate what we are seeing across the 50 US states in the contemporary period. Looking at the US through a comparative lens, each of these approaches holds out significant promise when considering the distinctive characteristics of the US political system. On the one hand, as one of the most decentralized federal systems in the world, the US might be considered a very strong candidate for the adoption of a unilevel approach, as Mickey has demonstrated. While the unilevel approach often has to make unrealistic assumptions about subnational units as freestanding, autonomous jurisdictions largely independent from the politics that unfold at the national level, these assumptions are arguably on safer ground in the context of such a radically decentralized political system.

Many of the features discussed above, including the absence of a reliable mechanism of federal intervention and extensive control by the states over municipalities through preemption, help to insulate and empower state-level autocratic actors. As we saw in the past, the highly decentralizing effects of Supreme Court decisions in the 1870s protected Southern states from the terms of the Thirteenth and Fourteenth Amendments and gave Southern political elites tremendous leeway and discretion to design and implement SURs to their liking. Other less political factors like the sheer size of the country might support the plausibility of a unilevel approach and the decision to bracket theoretically the role of national-level factors in order to be able to isolate the effects of largely subnational causes.

On the other hand, several important features of the US political system that are either missing or much weaker in Latin America and other parts of the Global South underscore the utility of a multilevel approach, as Gibson and Grumbach have shown. One could argue that the credibility of the unilevel approach is undermined by three critical factors that likely tighten the connection between levels of government in the US, and which therefore speak to the need for a more multilevel approach to theorization. First is the strength of the central state in a Weberian sense as an actor that has fully penetrated the national territory and monopolized the use of force vis-à-vis subnational state challengers. The extensive presence of the US federal bureaucracy, deconcentrated into all 50 states, exceeds what we see in Latin America’s federations and likely makes it harder for would-be subnational autocrats to escape federal influence. Second is the nature of the party system in the US, including the presence of two political parties that compete throughout the national territory and are far less decentralized than they were in the past, as well as the absence in the US of purely subnational parties, which are common in other federal democracies. Third is a set of economic factors including very high levels of interstate trade, a federal highway system, and a quite integrated national market. We know from the literature on Latin America that SURs flourish when autocratic governors can preside over closed economies that are largely insulated from national market pressures (Behrend 2016).

Conclusion

The primary goal of this article is to offer a new perspective about subnational democratic erosion in the US. Rather than compare what is happening today in the US to past patterns of erosion in this same country, we offer a new, comparative perspective that draws on the abundant scholarship on SURs in Latin America (and the Global South). As we show in this article, much can be learned from that literature to better assess the current process of subnational democratic backsliding in the US.

On a descriptive level, we find that despite a few common traits, including persistent attempts to undermine political opponents and erode checks and balances, on several other key dimensions the institutional landscape for the emergence of SURs should be considered more favorable in the US than it is Latin America’s three federations. First, voter participation is more restricted in the contemporary US as compared to Latin America, where we do not see concerted attempts to suppress electoral participation. Second, the parties that have built SURs in Latin America occupy a wider range of ideological positions than is the case in the US, which makes it harder for them to coordinate and scale up their efforts. Third, institutional mechanisms to erode SURs from above and from below are far less robust in the US than in Latin America. The prerogative of federal intervention, which has been a tool to prevent subnational democratic erosion, is not enshrined in the US Constitution. The opposite is true in Latin America’s federations, where this prerogative has been used to protect subnational democracy. Similarly, municipal autonomy is limited in the US but more extensive in Latin America. Studies have shown that subnational erosion can be prevented when municipalities, often ruled by the opposition, are able to stand up and defend democracy from below.

On a theoretical level, and considering what we have learned from Latin America, we argue that a multilevel rather than a unilevel approach is likely to be necessary to illuminate the phenomenon of subnational democratic erosion as it unfolds in the US. As noted above, the highly deconcentrated federal bureaucracy and the nature of the US national party system probably make it even harder for subnational autocrats to consolidate and entrench themselves in power without support from actors at higher levels of the political system. Unilevel approaches are often hard pressed to effectively bracket the influence of the national government, and this assumption seems harder to sustain in the US as compared to Latin American federations.

In addition to demonstrating that a comparative analysis of subnational democratic erosion in the Americas is descriptively and theoretically illuminating, we conclude with two related questions for future research on this issue. First, most of the literature on national democratic backsliding has been silent regarding space and territory. ⁷ What, if any, territorial strategies have national democratic backsliders adopted from above, including party building below the national level? How have Donald Trump, Jair Bolsonaro, Andrés Manuel López Obrador and, more recently, Javier Milei understood the role of subnational governments vis-à-vis the national democratic regime, and how have they fended off challenges posed by democratic governors and mayors in large cities? Second, from below, do subnational democratic backsliders seek to influence national-level politics and/or achieve national democratic erosion? If so, do subnational autocrats coordinate among themselves, and through which institutions or noninstitutional mechanisms? We hope that this article can help to expand the research agenda on democratic backsliding in several ways. First, by initiating a new comparative research agenda on subnational democratic backsliding in the Americas; second, by incorporating territoriality into the study of national democratic backsliding; and third, by encouraging discussion of which theoretical approaches are most likely to advance the study of SURs in federal countries.