Annual Review of Political Science Political Control

Political leaders use different tactics to ensure widespread compliance with state policies and to minimize resistance. Scholarship tends to treat different tactics individually, suggesting fundamental dissimilarities in underpinning logic and goals. We introduce political control as a concept that unifies these different tactics within a single framework and demonstrate the analytical utility of considering seemingly disparate strategies in conversation rather than in isolation. We synthesize a growing recent literature on political control, including innovative approaches to repression as well as studies of indoctrination, distribution, and infiltration. We argue that tactics of political control can be understood to vary along two primary dimensions: the level of violence and the materiality of state actions. We highlight recent inquiry into the downstream effects of political control. We conclude with a call for more research on political control that considers combinations of different tactics, across regime types, in a world where tolerance of violent repression is diminishing.


INTRODUCTION
Political control is at the heart of what states do. A functioning state requires that individuals pay their taxes, obey the law, and maintain the peace. How do rulers ensure that citizens comply with these state objectives? What tactics do political leaders use to minimize resistance, especially in weak democracies and autocracies? At one end of the spectrum of strategies of political control, states use violent repression to force citizens to obey. In the twentieth century, authoritarian regimes across the globe imprisoned, tortured, killed, and more generally terrorized their political opponents. However, political scientists have long recognized that even the most brutal states do not rely on violent repression alone, and more recent work suggests that states' use of overt, violent repression has been on the decline for decades (Guriev & Treisman 2019). At the other end of the spectrum is the use of information to manipulate citizens in an indirect and less violent manner.
Between the two poles of hard repression and information manipulation exists a broad range of strategies of political control that rulers use to shape the behavior of the citizenry. In this review, we survey the growing literature on political control. We define political control as tactics engineered by political leaders to ensure widespread compliance with state policies and to minimize political resistance. We conceptualize political control as an umbrella concept that brings together seemingly disparate state behaviors underpinned by a similar logic and driven by similar goals, in an effort to expand the scope of our collective attention beyond repression to other behaviors that might otherwise be overlooked but are essential to regime maintenance and state stability. We argue that tactics of political control vary along two primary dimensions. Rulers coerce compliance using violent and nonviolent tactics, ranging from repression to social sanctions. In addition, rulers use material and nonmaterial rewards, ranging from supplying state benefits to indoctrination, to encourage compliance.
We ground our review in concrete examples of political control to demonstrate the utility of considering disparate tactics in conversation and comparison with each other. While rulers can use many different methods of political control, we focus on four that have been the subject of scholarly analysis. First, we examine and summarize the large literature on violent repression. Next, we summarize how leaders use indoctrination to induce societal compliance. Third, we examine the use of welfare programs to co-opt and coerce compliance with the state. Finally, we examine how states infiltrate organizations in order to penetrate society and apply social pressure on citizens.
After our discussion of individual tactics of political control, we review research on their downstream unintended effects. Society often pushes back against state attempts at political control, regardless of the specific tactic that is used. Indeed, while backlash mobilization against violent repression is perhaps the best-studied example of this phenomenon, we draw on literature on downstream effects of coercive distribution and infiltration to highlight some of the costs that leaders encounter when they attempt to control society via any of the tactics that we describe.
This review makes two main contributions. Our core contribution is to show how moving beyond the study of repression-in a world where overt, violent repression is frowned on-reveals the ways in which leaders still use coercive tactics to control society. Recent research shows that the number of political killings and the number of political prisoners have been steadily declining (Fariss 2014(Fariss , 2019Guriev & Treisman 2019). 1 Yet, even if states have become gradually less violent, they still use a variety of tactics to ensure compliance and reduce political resistance.
In our review, we seek to demonstrate the utility of comparing seemingly disparate tactics of political control.
Our second contribution is to expand the growing literature on the state from discussions of capacity to also consider intent. Conceptions and measures of state capacity have tended to focus on the conditions under which states accomplish what they set out to do with only an agnostic examination of the tasks to which states apply that capacity. Political control is, instead, related to the concept of despotic power put forward by Mann (1984). Fully understanding the state requires both an understanding of its ability to shape society and an understanding of the ultimate manifestation of society that the state envisions. Put simply, the reason why a state implements a policy is analytically distinct from why that policy succeeds or fails, and it deserves its own study.
While tactics of political control are used by the full span of regime types, the most intrusive forms are concentrated in autocracies, where high levels of state control over society are necessary for regimes to stay in power. This review reflects this observation, and our focus on intent helps account for it. Most of the studies we highlight draw empirically from authoritarian regimes. While states governed by different types of regimes can and do rely on all methods of political control that we highlight, these tactics are more extreme-and thus easier to empirically observeunder authoritarianism.
The article concludes with a call for more research on political control. First, comparative studies must continue to unpack the intentions driving state actions and those elites who run them. While recent years have seen a surge of literature on the ability of states to carry out their core roles of governing, we encourage scholars to expand the scope of inquiry and to think critically about the purpose and motivations behind these actions as well as their unintended consequences. We also call for additional research that considers the combination and substitution of different strategies of political control, particularly in a world where overt repression may be waning. Finally, we call for additional research on political control that blurs boundaries of regime type and examines these tactics in democracies.

WHAT IS POLITICAL CONTROL?
Political control is any tactic through which the state seeks to gain compliance from society. Societal compliance is necessary for the state in its efforts to maintain order, enforce laws, collect revenue, implement policies, and minimize political resistance. That compliance may be subconscious, quasi-voluntary (Levi 1988), or achieved through force, but ultimately, political control entails state command over society to reshape society in ways that make it more legible and pliant to state directives. 2 The concept of political control is distinct from common understandings of state capacity. Mann (1984, p. 189) defines state capacity as the "ability to penetrate society and implement logistically political decisions through the realm." Put differently, state capacity is "the ability of state leaders to use the agencies of the state to get people in the society to do what they want them to do" (Migdal 1988, p. iii). One particularly illustrative example of state capacity is taxation. States that can effectively collect taxes are said to have strong capacity; those that cannot are weak. Political control, however, focuses not on the state's ability to impart change but instead on the To categorize tactics of political control, we think of variation on two dimensions. First, political control varies by the amount of violence used. When a strategy is more violent, it is more direct and overt: Intended targets know they are being targeted and are aware of the threat of force that might result if they do not change their behavior. When a strategy is nonviolent, it can be indirect and covert: Targets are sometimes, but not always, unaware that they are being controlled. Second, political control varies by the degree to which the state uses material and nonmaterial threats and inducements. Material inducements relate to tangible outcomes, including bodily integrity, money, and goods, while nonmaterial inducements are intangible and include social sanctions, worldviews, communities, a sense of belonging, and emotions.
Though political control happens across regime types, its usage in democratic and authoritarian contexts diverges in systematic and predictable ways. First, political control is used for different ends across regime types. In more democratic regimes, political control is meant to achieve societal compliance for the current leadership and all others to come. In more authoritarian regimes, however, political control is often in service of the survival of the ruling elite: The maintenance of law and order and the implementation of policy help current leadership stay in power. In addition, while states governed by both democratic and authoritarian regimes benefit from making society more compliant, regime types tend to vary in how they achieve compliance and what means of political control they favor. Society is often a more willing participant in political control under democracy. In choosing elected representatives, citizens under democracy help shape the contours of political control tactics and thus are often more willing to cooperate than citizens under less open regimes. In contrast, violent tactics tend to be concentrated under authoritarian regimes, where political control is often carried out by the regime's bureaucracy and security forces.

TACTICS OF POLITICAL CONTROL
In this section, we highlight four different tactics of political control emerging from recent literature: repression (both direct and by extension), indoctrination, coercive distribution, and infiltration ( Table 1). Our goal is not to describe all tactics of political control or to provide an exhaustive review of related literature, but rather to put concepts that are often treated separately into direct conversation, as strategies of the same type with differences of degree rather than kind.

Repression
The tactic of political control that has arguably received the most attention is repression. Repression refers to state behaviors targeting actors who challenge state beliefs, institutions, and actions (Davenport 2007). Repression comprises actual or threatened physical violence, including arrests, beatings, assassinations, torture, disappearances, mass killings, and forced exile (Davenport 2015, p. 29). The goal of the state in using repression is to force compliance by demobilizing or annihilating opposition actors.
The most straightforward type of repression is direct: A state targets a political actor for oppositional behavior. In direct repression, the violence is obvious, and the target is aware of the harm caused to their life and livelihood. A state need not engage in constant or substantial repression for this tactic to bear fruit. In fact, repression that is unpredictable-often due to weak state capacity-can still compel political control and undermine organized resistance to the regime precisely because it is applied arbitrarily (Tapscott 2021). While states mostly repress those living within their territorial jurisdiction, states also repress opposition outside of their borders. Transnational repression has increased over time in conjunction with global increases in migration, which in no small way has been exacerbated by domestic repression, as regimes seek to limit the influence of opposition actors from abroad (Moss 2016, Tsourapas 2020).
In contrast to direct and overt repression, other types of repression rely on the threat of violence. These repertoires include acts of harassment, intimidation, and administrative or legal blockages. States set up legal and institutional red lines that then lawfully permit them to use violence when these boundaries are crossed. Scholars refer to this less direct, less physical, and less violent form of repression in various ways-as "low intensity" (Levitsky & Way 2010), "institutional" (Koopmans 1997), or "covert" (Davenport 2005)-but the underlying logic remains comparable to that of direct repression: the threat (in contrast to the actual use) of force to induce compliance with the state.
Most states rely on some form of divided structures of contestation in an effort to economize when using repression (Lust-Okar 2005). However, the targets of repression vary widely across time and space. Repression can target either rival elites or the masses. The target of repression is a function of the interaction between the ability of the state to collect high-quality intelligence (Greitens 2016) and leaders' perceptions (Wintrobe 2000, Brownlee 2007, Svolik 2012 of which groups constitute a viable threat. In addition, repression varies in how widespread it is within elites or the masses and whether it affects all groups, only some, or one (Nugent 2020a). In a pattern similar to Kalyvas's (2006) typology of civil war violence, repression may be indiscriminate within groups; it may be aimed at wide geographic areas, large opposition groups, diffuse networks, or broad identity groups without consideration to the actual behaviors or opinions of any one elite or individual. Alternatively, repression can be discriminate and punish specific individuals. For instance, purges are frequently used against the incumbent regime's elite rivals (Sudduth 2017, Goldring & Matthews 2021, Woldense 2022, and preemptive societal repression is reserved for the individuals or groups that are expected to mobilize against the state, often in anticipation of important focal moments (Truex 2019).
How does repression succeed in creating political control? Much research has found that states repress individuals and groups who are at the forefront of resisting the state. This shifts political opportunity structures, raising the costs of opposition behaviors and ultimately forcing actors underground or exterminating them (Davenport 2015, Fu 2018, Amat 2021. Repression can also exert political control by generating fear. For individuals in Syria, the regime of President Bashar al-Assad generates a "silencing fear" that induces submission through the threat of punishment (Pearlman 2016). When looking at Ukraine's Holodomor, during which nearly 3 million people starved to death, Rozenas & Zhukov (2019) find that areas hardest hit by this famine were those in which locals were most unwilling to support groups opposed to Moscow. Recent work directly manipulates fear to test its role experimentally: Young (2019) finds that when the emotion of fear is primed, Zimbabwean dissidents are less willing to engage in antiregime activity because they are more pessimistic about the likelihood of success. In turn, fear of individual harm can induce collective compliance as communities self-police for fear of collective retribution (Rozenas 2020).
Repression as a strategy of political control typically relies on material inducement; the use or threat of state violence decreases the quality of life and the bodily integrity of those targeted. However, different kinds of repression may rely on immaterial inducements. While it is not necessarily a new tactic of political control, scholars have increasingly documented and theorized "coercionby-proxy" (Tsourapas 2020, pp. 627-28) or repression by extension. In this form of repression, states repress individuals or groups in an attempt to control the behavior of other individuals or groups who are beyond the reach of the state because they are either abroad or in hiding.
Repression by extension is most clear in instances when the state represses the immediate or extended family of dissidents. This tactic uses a similar approach to the use of violence but induces compliance by preying on the emotions of the opposition. In Tunisia, under President Zine el-Abidine Ben Ali, the state repressed the wives of activists in exile and arrested the brothers of those in hiding to force members of the opposition into making themselves visible. In contemporary Egypt, the state subjects the family of exiled opposition to arbitrary detention and false imprisonment for their relatives' activism. A similar logic underscores what Deng & O'Brien (2013) call "relational repression" in China, where officials pressure those with social ties to activists to exert personal influence, demobilizing the activists through immaterial inducements of interpersonal connection, accompanied by the threat of material inducements for those who refuse to comply.
Repression by extension can also induce political control at a group level. When the state represses groups who stand in opposition to other groups (and the state), the state-and the politicians at its helm-may gain legitimacy in the eyes of, and popular support from, the nonrepressed group (Wilkinson 2006). This amounts to a repressive version of "the enemy of one's enemy is one's friend." Lachapelle (2021) finds evidence of this dynamic in Egypt, where the regime of President Abdel Fattah el-Sisi consolidated power and built popular support among non-Islamists by repressing the Islamist Muslim Brotherhood. The state's use of repression showed the regime's "type" and implicitly signaled to non-Islamists that the president was a copartisan who would look out for their interests. In repression by extension, states still induce social compliance through violent tactics. But the change in behavior is less a product of the bodily harm felt by the repression's subject (as in direct repression) and more due to the reactions of others when they witness or hear about said repression.

Indoctrination
Scholars have often focused on the positive effects of public education in creating citizens who possess the relevant skills, knowledge, and worldview for becoming members of a national community. At the same time, education is at its core a process of indoctrination and socialization, and it can be used by states to increase compliance and control.
Recent research challenges the idea that mass education increases with democracy and the expansion of other rights and liberties. The conventional wisdom has been that the provision of education redistributes political power from rich to poor and is part of the process of democratization (Stasavage 2005). Yet, recent research finds that it is authoritarian regimes that tend to expand mass education; on average, states expand mass education nearly a century before the first transition to democracy (Paglayan 2021) and often do so when they are recovering from civil conflict (Paglayan 2022).
The process of indoctrination that occurs in public education systems is a noncoercive strategy; no actual or threatened violence is employed. In addition, it can rely on both material and immaterial inducements to targets. For example, indoctrination may promise material inducements such as increased wages and better standards of living as states seek to create or increase industrialization (Gellner 2008) or militarization (Darden & Mylonas 2016). In terms of immaterial inducements, indoctrination may emphasize community acceptance as states seek to forge a national identity, create allegiance to a national project (Fouka 2019), or promote domestic order through loyalty to the state (Darden & Grzymala-Busse 2006, Cantoni et al. 2017, Pop-Eleches & Tucker 2017, Alesina et al. 2021, Paglayan 2021. Indoctrination need not end with school. Propaganda through print and social media may, for instance, paint a picture of the state's proficiency in governing (Rozenas & Stukal 2019), incite anger at the state's rival powers (Mattingly & Yao 2022), or distract citizens from the government's failings so as to prevent collective action (King et al. 2017). Censorship can similarly shape citizen opinions by eliminating from public discourse information or opinions that may make compliance with doctrines of political control more difficult. For example, states censor to prevent collective action (King et al. 2013) and limit societal discontent (Lorentzen 2014). And indoctrination can make bureaucrats better agents of the regime, more consistently implementing regime elites' preferences, even if the indoctrination content does not change their ideological positions (de Juan et al. 2021).
Further, censorship of some societal groups may meet the policy preferences of other societal groups and compel their continued support. In Chile under President Augusto Pinochet, censorship of popular media targeted not only antiregime content but also content that was seen as inappropriate for society. The goal, according to Esberg (2020), was to increase and maintain the support of Chile's Christian right, an important base of popular support for the regime. This research has analogs to repression by proxy as described above. Regimes may command the state to engage in a tactic of political control-such as repression or censorship-to build support among a nonaffected group. More generally, this parallel speaks to our broader point: Placing tactics of political control in conversation with each other allows us to see how they are simply different methods that leaders use to achieve similar ends.
Indoctrination may also go through a myriad of other avenues. Consider the courts. Shen-Bayh (2018, 2022) examines reliance on political trials against coup plotters. Since a range of options is available to repress elites (Esberg 2021), under what conditions can we expect punishment through the legal system? Shen-Bayh argues that the purpose of a political trial is not the punishment inflicted on the coup plotter but instead the narrative that the state writes through the public ceremony of the trial. In particular, trials help authoritarian leadership reframe the messy reality of political conflict into a simpler narrative of right and wrong. Performing this narrative in court before an audience helps generate shared beliefs about acceptable (and unacceptable) behavior, which help dissuade future challengers and enforce submission to power.
Recent literature suggests several ways in which indoctrination leads to societal control. For one, indoctrination can persuade. For example, reforms to political education in Chinese high schools increased trust in officials, made students believe China was more democratic, and made them more skeptical of markets (Cantoni et al. 2017). For another, indoctrination can be used to signal state strength. When autocracies produce propaganda that makes outrageous claims-for example about the health and intelligence of leaders like Assad, Stalin, or Mao-it may not make individuals genuinely more supportive of the state or leader. Instead, this propaganda forces people to behave as if they support the regime, and in doing so creates "guidelines for acceptable speech and behavior" and "induces complicity by creating practices in which citizens themselves are 'accomplices'" (Wedeen 2015, p. 6). In China, recent evidence shows that exposure to hard propaganda does not persuade people to support the state but does make them more politically quiescent (Huang 2015(Huang , 2018. Spectacles like televised show trials can be used to threaten dissidents, outline the boundaries of acceptable political behavior, and project the image of a benevolent state (Fu 2020).
In addition, indoctrination can inflame emotion. Indoctrination can heighten feelings of fear or anger, which in turn shape how citizens register additional information or how they act upon information. For instance, even if propaganda is known not to be credible, it may induce sufficient fear among dissidents that they refrain from acting against the regime (Little 2017. At other times, indoctrination works by provoking nationalist anger toward a foreign adversary, as seen in the Chinese Communist Party's use of propaganda to inflame anger against Japan (Mattingly & Yao 2022).

Coercive Distribution
States can distribute welfare and state resources in a manner that controls citizens and elites alike. Land reform, poverty alleviation, public health, housing, education, and employment programs make citizens dependent on the state and allow rulers to coerce citizens into complying with their demands. Albertus et al. (2018) call this phenomenon "coercive distribution." They argue that "distributive programs are an advantageous way for young authoritarian regimes to confront both elite and mass threats" (p. 1). At the mass level, "public service provision can bind mass populations to regimes by making the regime, and the state apparatus it controls, the center of citizens' 'strategies of survival'" (p. 10). When citizens are dependent on the state for land, housing, and other services, their dependence gives the state leverage over citizens, and "a regime's threats become more credible and more ominous-and its replacement less thinkable" (p. 10). This strategy "undercuts collective action in a way that overt repression typically does not" (p. 12).
In newly founded authoritarian regimes, Albertus and colleagues show how land distribution can weaken the power base of political rivals in the landlord class (Albertus 2015, Albertus et al. 2018. In Peru, "General Velasco seized power in a military coup in 1968 and quickly decreed an agrarian reform law in 1969 that aimed squarely at landed elite rivals to the regime" (Albertus et al. 2018, p. 24). Similarly, Castro in Cuba and Mao in China used land redistribution in the early years of Communist rule not just to fulfill promises made during their respective revolutions but to wipe out a landlord class that might otherwise have challenged their power. Even Mao's rivals (the Nationalists under Chiang Kai-Shek) began their rule over Taiwan by starting a program of land reform designed to undermine an indigenous landlord class that had long been an important element of the island's governance (Read 2012, Matsuzaki 2019a. In more established authoritarian regimes, state control over employment can also help to create a quiescent middle class with autocratic tendencies. Rosenfeld (2017Rosenfeld ( , 2020Rosenfeld ( , 2021 shows that state-sector jobs make individuals less likely to support democratic reforms or protest across numerous post-Communist countries, most notably Russia. Concerns over one's welfare may build middle-class support for autocratic parties by shifting material incentives. It is not necessarily the case that members of the middle class will lose their jobs if they do not vote for autocratic parties such as United Russia, but their secure positions in the state sector make them more inclined to support the incumbent party. This dynamic is less coercive than the logic of coercive distribution highlighted by Albertus et al. (2018). As Rosenfeld (2020, p. 51) notes, "Material benefits are used not only to coerce, but also to persuade." State welfare programs can also be tools for monitoring and defusing dissent. Pan (2020) shows that China's main poverty alleviation program, called dibao, not only provides services to poor families but is used by the state to identify and preempt antiregime protest. However, the program also generates significant backlash among nonrecipients; Pan (2020, pp. 152-62) finds that locales with large numbers of dibao recipients on balance experience more protest and have lower state legitimacy. Thomson (2019) finds a similar dynamic. He builds on early insights by Bates (1981) to examine the conditions under which states extract agricultural resources from the countryside. When the dominant popular threat is urban unrest, states lower urban food prices at the expense of rural incomes despite ruralites' potential for countermobilization. Taken together, these authors show that the use of welfare for political control can come with a significant cost: backlash among groups excluded from the programs. While the notion of backlash mobilization is familiar to scholars of repression, we see how a seemingly dissimilar tactic can lead to similar unintended outcomes when we expand our understanding of political control.
It is not just consolidated autocracies like China and Russia that use welfare programs for political control. Mares & Young (2019) show how, in democracies such as Hungary and Romania, local politicians trade access to welfare programs for votes (see Hicken & Nathan 2020 for a more general review of clientelism). For example, in a small Hungarian town, a mayor exploits his control over a "workfare" program that provides steady income in return for working on public projects by threatening "that current beneficiaries will lose access to the program and hence, to an important stream of future income, if they fail to support his preferred candidate" (Mares & Young 2019, p. 2). The dynamic these authors highlight is more coercive than the "autocratic middle class" thesis of Rosenfeld (2020), since it involves an explicit threat of losing welfare benefits if the recipient does not vote the right way.
For another example of coercive distribution, again consider land. Research on the property rights gap finds that many states allocate land to peasant farmers but withhold the associated land rights (Mattingly 2020a, Albertus 2021, Hassan & Klaus 2021. The lack of property rights is not a result of weak state capacity; a state that has the ability to expropriate and reallocate land likely has the capacity to grant the associated property rights (Albertus 2021). Instead, states withhold rights to keep farmers land insecure and thus maintain a tool by which to compel specific behaviors and maintain rural acquiescence. While reliance on the property rights gap holds across regime types, which societal groups the state manipulates through the closing of the property rights gap changes with regime openness: Whereas authoritarian regimes are more willing to bestow property rights among their core, over whom they hold other levers to compel societal control, electoral regimes are more likely to close the gap among swing voters in a myopic attempt to win an upcoming election (Hassan & Klaus 2021).
The tactic of coercive distribution is linked to the concept of co-optation. Broadly defined, co-optation is a nonviolent process through which states induce compliance through the incorporation of groups that emerge autonomous and separate from the state. In more concrete terms, co-optation can be defined as "the intentional extension of benefits to potential challengers to the regime in exchange for their loyalty" (Frantz & Kendall-Taylor 2014, p. 333). This type of political control is nonviolent and garners compliance through the exchange of a tangible good; the co-opted cooperate, and in exchange the state grants policy concessions or economic rents (Gandhi & Przeworski 2006).
Co-optation exists in democracies but is arguably more deliberate in nondemocratic regimes, where states co-opt groups like opposition parties, civil society, and rival elites to aid in the state's survival and stability. One example of co-optation is the use of parliaments in nondemocratic regimes. Authoritarian regimes co-opt opposition to participate in unfree and unfair elections in exchange for access to state spoils for constituents as well as personal benefits like business opportunities and increased income (Magaloni 2006, Blaydes 2010, Opalo 2018. When the opposition wins seats in elections and serves in government, it participates in a legislature that is essentially a rubber stamp (Brancati 2014, Moustafa 2014) and in doing so lends legitimacy to the state. Here, compliance manifests as cooperation: the co-opted opposition participate in supporting the state by overwhelmingly voting in favor of the state's policy agenda (Tavana & York 2020). Co-optation of rival elites works similarly, though here compliance manifests not only as cooperation but also as inaction. For example, co-opted rivals may decline to levy criticism or launch a coup against leaders of the state (Bove & Rivera 2015).

Infiltration
Another common tactic of political control is to infiltrate society using bureaucracies, local community elites, or civil society to collect information, mobilize society, and induce compliance with central directives. Infiltration may be most common in autocracies but is also used in weak democracies to demobilize the opposition and mobilize support.
Many socialist governments of the twentieth century created far-reaching organizations to monitor and control society at the grassroots level. The most infamous was the vast network of secret informants recruited by the Stasi in East Germany. However, many infiltrating organizations operated in the open. The Communist Party in Cuba created neighborhood-level organizations called Committees for the Defense of the Revolution (CDR) whose members served as the local eyes and ears of the state. When Prime Minister Fidel Castro announced the creation of the CDRs in 1960, he declared: "We're going to set up a system of revolutionary collective vigilance so that everybody will know everybody else on his block, what they do. . .what they believe in, what people they meet, what activities they participate in" (quoted in Read 2012, p. 250). Similarly, the sprawling structure of Tunisia's Democratic Constitutional Rally (RCD), the country's hegemonic ruling party prior to 2011, was intended as a means through which to collect intelligence on party members through reports and surveillance from other members. Instead of governing, the RCD was designed to infiltrate and surveil political opponents (Hibou 2011, pp. 87-93).
The Chinese Communist Party (CCP) is an especially prominent user of this tactic. Read (2012) demonstrates how the CCP uses residents' committees as the "nerve tips" (Read 2000) of the state and how these groups are central to party rule and political control. Mattingly (2020a) and Pan (2020) show how neighborhood administrators in rural and urban China alike remain essential for implementing policy and identifying dissent. The importance of this tool was highlighted during the coronavirus disease 2019 (COVID-19) epidemic, when neighborhood committees took the lead in monitoring and enforcing compliance with lockdown orders. Similar neighborhoodlevel groups were also an important part of colonial and Nationalist rule over Taiwan (Read 2012, Matsuzaki 2019a. Labor unions are another important tool for political control through infiltration. In Latin America, "state control of the working class. . .was achieved at least in part through the legalization and institutionalization of a labor movement sanctioned and regulated by the state" (Collier & Collier 1991, p. 3). For example, the Democratic Action Party used ties to union leaders in order to organize and control workers in Venezuela. Collier & Collier (1991, p. 167) argue that corporatist states can "repress. . .preexisting unions and replac[e] them with highly constrained, state-penetrated labor organizations that would avoid class conflict and instead 'harmonize' the interests of capital and labor." New research builds on the canonical contributions of Collier & Collier (1991) and furthers our understanding of corporatism for political control. In China today, labor unions are tightly controlled by the CCP and help the state identify and demobilize dissent (Friedman 2014;Fu 2017Fu , 2018. Similarly, in President Hosni Mubarak's Egypt, labor leaders were willing to work with the state to keep labor docile amid painful neoliberal economic reforms, and the union rank-and-file followed the direction of union leaders (Bishara 2018). The state then protected compliant union leaders to maintain their elite positions while hindering more radical union members from standing for union office. As a third example, Anderson (2016) finds that labor's ties to the Moroccan monarchy both limited the scope of demands to socioeconomic (as opposed to more political) reforms and confined the way in which labor made their demands, privileging intra-elite negotiations instead of disruptive demonstrations and strikes.
Relatedly, we have seen a reinvigoration of research into the conditions under which statecontrolled institutions may facilitate mobilization. In Russia, for instance, corporatist linkages between elites and unions have led to varying levels and types of protest (Robertson 2010). When and where regional elites have no other recourse to attract central attention, Russia sees elitesponsored mass mobilization to demand central government transfers. In recent years, as the opposition to President Vladimir Putin has strengthened, the state has used its corporatist linkages to organize state-sponsored mobilization. Activists in Egypt push for policy concessions using the spaces and resources granted by state-controlled unions (Bishara 2018). In democratic Brazil, bureaucrats empower societal advocacy groups to help them mobilize and externally push for bureaucrats' preferred policy changes (Rich 2019).
In a similar vein, much of the rich research on African traditional chiefs explores the state's use of societal elites as intermediaries to infiltrate the communities they oversee. Historically, colonial powers and authoritarian regimes alike formalized chiefs' authority over their subjects (Bates 1981, Boone 2003, Riedl 2014. These intermediaries were expected to use their local societal clout, backed by the power of the state, to compel citizens to behave in the manner that the center dictated. And indeed, chiefs were able to force a range of behaviors, from the payment of agricultural taxes (Kasara 2007) to compliance with new education standards (Nathan 2021).
As such, the continued formal authority of chiefs during the continent's current electoral era seems anachronistic; these unelected local elites often serve for life with little means of accountability. Yet, precisely because their authority depends on state consent, states see value in continuing, and even further empowering, these intermediaries (Baldwin 2014). Unsurprisingly, research has found that chiefs continue to use their clout to persuade, and at times pressure, their subjects into compliance with broader state objectives (Acemoglu et al. 2014, Koter 2016 Kadt & Larreguy 2018, Nathan 2021).
As described above, infiltration tends to involve the capture of local elites. The state leverages organic social connections between local elites and citizens and pushes local elites to persuade citizens. There are, however, conditions under which well-embedded state officials can serve as effective infiltrators and similarly persuade subjects. In Kenya, from the colonial period to the current multi-party era, centrally appointed bureaucrats have compelled citizens in their jurisdiction to, for instance, provide labor for local development projects and stop the cultivation of forest land (Hassan 2020). In many cases, citizens comply not because of the threat of state repression but because of the social bonds that bureaucrats have or develop with citizens. Well-embedded bureaucrats, through either coethnicity with locals or long tenure in a jurisdiction, become the local societal elites that can elicit compliance through persuasion.
Civil society itself can become an instrument of infiltration and political control. The dominant view of civil society is that it stands as a bulwark against the state and is a force for democratization. Yet, civil society groups do not always stand in opposition to authoritarian regimes. For example, Jamal (2009, p. 24) examines the role of civil society groups in Palestine and the Arab world and shows how states can "co-opt associational leaders, urging them to adopt more sympathetic and government-supporting stances." She concludes that nongovernmental groups "will reproduce elements of the political context in which they exist. . . . Where associational contexts are dominated by state-centralized, patron-client tendencies, then associations, too, become sites for the potential replication of these vertical ties" ( Jamal 2009, p. 20).
Indeed, authoritarian regimes generally only allow nonstate groups to exist when they help the regime survive or at least do not interfere with it. For example, in rural China, temple and lineage associations can bind local officials and citizens in relationships of mutual obligation, which can provide social incentives for officials to provide better public services like roads and schools (Tsai 2007). Yet, these same groups can also give local officials leverage over society. These groups help officials co-opt local elites whose cooperation is needed to enforce unpopular policies around taxation, family planning, and land (Mattingly 2016(Mattingly , 2020a. Similarly, nongovernmental organizations (NGOs) help the Chinese state collect feedback and information about grassroots preferences, which help the state endure (Teets 2013(Teets , 2014. Even NGOs whose explicit mission is democratization can help authoritarian regimes consolidate control over society. Bush (2015, p. 5) demonstrates that "many 'democracy assistance' efforts today in fact do not confront dictators" but instead support NGOs with "tame" programs that "refrain from directly confronting dictators" and "help organizations promote their survival." When organizational survival depends on consent from the leaders of authoritarian states, even nominally pro-democracy civil society organizations may in effect help authoritarian regimes control society and defuse dissent.
When do states invest in infiltration? Authoritarian regimes condition their infiltration strategies on the structure and strength of existing social groups. Where a strong civil society already exists, infiltration is more readily and successfully resisted, and thus states invest less. The Patriot Priest program in communist Poland, which recruited regime sympathizers from among the priesthood in the Catholic Church, provides an example (Nalepa & Pop-Eleches 2021). The Communist Party targeted areas for heavy infiltration where the church was weaker prior to the communist takeover. Similarly, in rural China, the CCP invests in neighborhood-level monitoring organizations where existing social groups, like clans and religious groups, are weakest (Mattingly 2020a). When authoritarian regimes want to recruit informants among regime opponents, they have to pay a steep price. Piotrowska (2020) shows that in East Germany, the Stasi had to pay significantly more for informants in towns that had freer media environments.
Where the state cannot readily identify potential infiltrators, it creates them. Across West Africa, for instance, colonial and postcolonial states alike empowered local elites with varying levels of formal state authority to create intermediaries (Boone 2003). Nathan (2021) richly details the experience of Northern Ghana where the colonial state "invented" traditional chiefs among acephalous (nonhierarchical) societies. Over time, doing so provided the state with a socially embedded and legitimate point-person to implement policy and control subjects. Some recent work on decentralization leads to a similar conclusion. States engage in this reform, in part, because it creates new local-level bureaucratic positions through which the center can better monitor local behavior (Lewis 2014), compel regime support (Hassan & Sheely 2017), or extract resources from local areas (Kosec & Mogues 2020).
The most consistent finding from the literature is that infiltration is most effective when states penetrate denser networks of agents. For example, Forrat (2018) focuses on the role of teachers as important grassroots agents of the state in Russia who engage in electoral agitation and fraud to assist the United Russia party. Since polling places are often in schools, teachers play a direct role in managing and counting the vote, and the author finds that higher density of teachers is correlated with more electoral support for Putin. Similarly, Mattingly (2020a,b) finds that locales that invested in denser neighborhood monitoring organizations had higher compliance with China's contemporary family planning policy. Nalepa & Pop-Eleches (2021) also find that in Poland, when the state recruited denser networks of patriotic priests, it increased local support for the regime.
How does infiltration work in creating compliance? The literature generally highlights the effectiveness of local operators who are deeply embedded in local society and outlines three mechanisms: persuasion, information, and coercion. Street-level bureaucrats with tight connections to the local community may be able to leverage their positions of trust and authority to persuade individuals to comply with and support the state (Hassan 2020, Nalepa & Pop-Eleches 2021. They may be able to use their position in social networks to acquire better information about social compliance and more effectively target the noncompliant for sanctions. Or they can use their social ties to coerce individuals using social pressure, social sanctions, or their control over local resources (Deng & O'Brien 2013, Forrat 2018, Mattingly 2020a, Pan 2020, Nathan 2021.

UNINTENDED EFFECTS OF POLITICAL CONTROL
In many cases, the tactics states use for political control have helped those states achieve their desired outcomes in creating societal compliance. But in addition, scholars have shown many downstream effects of control tactics that are unintended and unrelated to their primary purpose. We first examine the large and developed literature on backlash mobilization against state repression, and how this violent tactic may undermine leaders' goals of social control. We then review newer research which demonstrates that other tactics of political control may have similar unintended effects on society and that citizens push back against state attempts to control in ways that may undermine the very stability that those tactics hoped to achieve in the first place.

Backlash Mobilization Against State Repression
Most prominently, policies aimed at state control via repression can sometimes result in backlash collective action against the state. For instance, Roberts (2020) reviews the myriad ways in which netizens have shown resilience to online censorship and proved their willingness to circumvent these state controls. Surveys with politically active Zimbabweans before, during, and after the contested 2018 elections suggest that repression mobilizes short-term opposition (LeBas & Young 2021). Repression can also lead to an increase in antiregime attitudes that persists for decades. In China, for example, individuals exposed to violent repression during the Cultural Revolution had stronger antiregime attitudes some four decades later (Wang 2021).
At times, the backlash effect extends to those observing the repression or individuals connected to the victims of repression. For instance, though most Tunisians repressed by the Ben Ali regime were demobilized, their family and community members showed higher levels of democratic participation after the Arab Spring (Blackman et al. 2020). Children whose parents were exposed to more violence during China's Cultural Revolution have stronger antiregime attitudes today (Wang 2021). This relational backlash effect transmits to online networks as well. Well-known Saudi activists who were imprisoned were deterred from dissent after their release-that is, political control seems to have worked as intended on these high-profile cases. However, engaged followers of these imprisoned activists engaged in backlash: They increased their online criticism of the monarchy and demands for political and social change (Pan & Siegel 2020).
Precisely because research has shown evidence for any and all relationships between repression and mobilization (Davenport 2007), recent work seeks to identify the conditions under which state repression leads to backlash versus demobilization. Research has begun to explore the effects of network structure (Siegel 2011), emotions ( Jasper 2011;Pearlman 2013;Young 2019Young , 2020, organizational structure (Sullivan 2016), and preventative repression (Pierskalla 2010, Ritter & Conrad 2016. We believe that untangling the relationship between repression and backlash mobilization holds promise for future research.

Unintended Societal Effects of Political Control
Recent scholarship has begun to examine other unintended effects of repression and strategies of political control more broadly. Much of this work has focused on societal and economic effects of political control.
For instance, new work has shown that state policies meant to control the citizenry have downstream effects on societal relations. Nathan (2021) investigates the long-term consequences of the colonial invention of chieftaincy for important societal outcomes using a natural experiment. The British did not invent chiefs among all acephalous groups. In turn, he finds that villages inhabited by groups in which chieftaincy was invented have higher levels of socioeconomic inequality today. Thacil (2020) looks at the effects of repression on migrants in India. Using ethnography and survey experiments, he finds evidence of migrants' increased willingness to economically and politically cooperate after police repression.
Further, political control's effects on society may feed back into state capacity. Charnysh (2019) examines the effects of the state resettlement of 5 million people into present-day Poland in the wake of World War II, when areas abandoned by Germans were resettled by other Europeans. Leveraging the arbitrary nature of relocation procedures and the resulting variation in villagelevel heterogeneity, she finds that more homogeneous resettled villages were better at community provision. In contrast, more heterogeneous villages instead came to rely more on state provision. However, the need of heterogeneous villages to rely on state provision has resulted in higher levels of local tax capacity and, ultimately, the strengthening of the state in these regions.
Political control that manifests as coercive distribution may have negative long-term political effects. Berman & Nugent (2020) find that long-term patterns of economic favoritism and marginalization under Ben Ali in Tunisia produced patterned regional heterogeneity in the attitudes and preferences linking voters with parties in later democratic elections. More specifically, attitudes associated with coercive distribution varied in their linkage with parties associated with the old regime and the old opposition. In addition, the coercive distribution of land can have negative economic effects. Albertus et al. (2020) link Peru's property rights gap to lower levels of educational attainment. One proposed mechanism for which they find evidence is that land reform beneficiaries wanted to make the most of this economic windfall, and so the farm labor of children became too valuable to allow for their schooling.
Research has begun to examine the mechanisms by which political control can lead to these unintended effects. One particularly interesting mechanism uncovered by recent work is identity. Repression, and the trauma it inflicts on targets, shapes political identities through psychological processes, and those identities are in turn reinforced by organizational and socialization processes. For example, those who were victimized during the Spanish Civil War and Franco dictatorship rejected a nationalist identity (Balcells 2012). The unintended consequences of repression on not only identity but also political polarization speak to the longevity and weight of these effects (Nugent 2020b). Identities conditioned by variation in repression contributed to authoritarian retrenchment in Egypt and continued democratic consolidation in Tunisia (Nugent 2020a). Blaydes (2018) demonstrates that Iraqi President Saddam Hussein and his regime used collective punishment of certain social groups. Over time, repression cemented sectarian divisions in society and created disastrous conflict along these lines after the American invasion destabilized the country in 2003.
The unintended consequences of repression continue to affect political outcomes through identities forged under trauma long after repression ceases. Fouka (2019) finds that German immigrant children who experienced German language restrictions in the United States during World War I showed lower levels of cultural assimilation as adults. Their behavior suggests that forms of political control outside repression-in this case, indoctrination-can similarly result in backlash effects.

CONCLUSION
We intend for political control to serve as an umbrella concept linking a number of tactics that states use to create and maintain societal compliance. Strategies of political control vary along two dimensions: whether they are violent or nonviolent, and whether they are material or nonmaterial ( Table 1). We reviewed new research on four political control strategies: repression, coercive distribution, indoctrination, and infiltration. We chose these topics because there is a sufficient literature on each and because they demonstrate variation in threat and inducement types. Taken together, these four strategies show how the concept of political control brings together topics that are currently treated separately but that, in fact, illuminate more about politics when considered in conversation with each other.
In concluding our review, we outline three areas for future research on political control. First, though many of the studies we discuss here isolate the causes and consequences of a single type of political control, the empirical record demonstrates that states rarely rely on only one tactic to maintain societal compliance. Additional research is needed to understand the relationship between different tactics, whether and how they are paired together, whether they function as substitutes, and whether different combinations of tactics have an interactive effect in creating and maintaining political control. For instance, Carter & Hassan (2021) show that whereas states invest in bureaucratic infiltration in zones of support, they turn to repression in opposition areas.
One particularly robust result in the literature is that the presence of institutions of co-optation is significantly correlated with a lower use of repression (Geddes 2005, Frantz & Kendall-Taylor 2014. We encourage scholars to investigate whether and how states combine different types of political control-and to do so beyond the standard repression-co-optation pairing. Important questions to answer include: Should we understand different tactics as substitutes, or as tactics employed in tandem, perhaps with predictable pairings? Is political control a bundle of strategies that states use in isolation, in combination, and at different times? How do strategies of political control ebb and flow over time? What predicts when states change the predominant political control strategy or combination of strategies? A second clear finding is that obvious, direct, and violent repression is rapidly becoming politically and economically costly for states, as political actors turn to digital opposition and the human rights movement goes global. As a result, contemporary states increasingly use tactics of political control that are nonviolent and rely on immaterial inducements. One example is the turn to "informational autocracy," in which regimes use a mix of propaganda and censorship to convince the public of their competence and superiority (Guriev & Treisman 2019). Scholars must continue to explore the nonviolent approaches and immaterial inducements that states use to induce compliance, comparing and contrasting these with more traditional aspects of political control. Less violent strategies using immaterial inducements are more difficult to quantify and thus harder to research, and they may also be less amenable to experimental approaches because they depend on internal cognitive processes. However, these characteristics make these tactics all the more important for scholars to understand, and researchers must be creative in using interpretive, ethnographic, and mixed-method approaches to identify and uncover evidence of these processes (Fu & Simmons 2021).
Finally, although we draw heavily on evidence from consolidating democracies and authoritarian regimes, our conceptualization of political control considers tactics to induce societal compliance across regime types. We encourage scholars to think beyond regime type dichotomies as they www.annualreviews.org • Political Control continue to explore political control. All states use political control in some way, regardless of the regime type governing their political system. The extent to which strategies differ in democracies and nondemocracies remains an open question ripe for careful comparative research inquiry. Davenport (2015) has demonstrated how logics of repression extend to consolidated democracies such as the United States. Additional research in this vein is needed to understand whether democracies intentionally co-opt and infiltrate opposition, distribute resources in discriminate ways, and indoctrinate citizens in the name of societal compliance, and whether they do so differently than autocracies.

DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.