What is the difference between civil associations and political associations? On March 13, 2021, the anniversary of the killing of Breonna Taylor following a “no-knock search warrant” by police in Louisville, Kentucky, multiple groups gathered in her memory in downtown Washington, DC. In emails made public by a lawsuit by the Brennan Center for Justice and Data for Black Lives, the DC Metropolitan Police Department monitored social media posts about these vigils. The police seemed to believe that one social media post, because it was shared by a user known to police and called for attendees to don “black bloc” clothing, posed a risk of “civil disobedience and/ or criminal activity” or even be the work of an “outside instigator” (Brennan Center for Justice 2024; see also Dyson et al. 2024). With its roots in a 1981 face-off between “German squatters and police evictors,” “black bloc” clothing entails dressing in all black along with masks, hats, and other anonymizing and protective gear, a style that has become popular among anarchist, libertarian-socialist, and anti-fascist protestors (Paulas 2017). Ironically, it is precisely the reasonable fear of surveillance, whether via biometric and AI-powered tools used by the police, or of “doxxing” by online mobs, that has made such anonymizing strategies popular.

Although “black bloc” protestors were particularly scrutinized by the authorities, the MPD had compiled a long list of other planned events and protests relating to racial justice during the summer of 2020, including religious organizations like Faith Works, whose flier emphasized it would be a “peaceful demonstration” with song and prayer (Dyson et al. 2024). What is a “public vigil” if not a voluntary and sacred moment of communal mourning and support? To the extent the memorials for Taylor were “political” gatherings, they were christened so by the state violence that brought such grief into the world. Nevertheless, the gaze of the state made no distinction between civil and political organizations.

Then why do celebrants of civil society as the bedrock of democracy so often draw a distinction between civil and political associations? Political scientist Robert Putnam’s Making Democracy Work: Civic Traditions in Modern Italy (1993) famously argued that the success of democracy depended on robust networks of social capital, developed through primarily apolitical civic associations, as opposed to participation in political organizations like parties. Putnam compared his theory to Alexis de Tocqueville’s emphasis on how “intellectual and moral associations” shaped the moral character of early Americans (1993, 90, 221n28). Putnam’s emphasis on social culture and civil society organizations resembles canonical thinkers in the classical liberal tradition who also celebrated Tocqueville’s emphasis on customs and civil associations (Hayek [1944] 2007; Cornuelle [1965] 2017; V. Ostrom 1998).

Nevertheless, the relationship between civil society and democracy cannot be understood without attending to the broader institutional context. In some cases, as Sheri Berman argued in an early response to Putnam, “a robust civil society may not produce beneficial effects, but rather may signal—and hasten—a democratic regime’s degeneration” (Berman 1997a, 562). The difference lies in the responsiveness of a society’s governing institutions and the degree to which citizens consider their institutions to be effective and legitimate. A disillusioned populace fed up with formal political parties and institutions might increasingly participate in civil society organizations. However, those seemingly apolitical organizations can be used by illiberal and antidemocratic actors to mobilize support for the subversion of institutions. Berman cited the flourishing associational life in the Weimar Republic that was exploited by Nazis in Germany as one such example (Berman 1997b).

As Berman acknowledged, however, Tocqueville valued both civil and political associations as necessary for democracy. By recontextualizing Tocqueville’s theory of civil society within his comments on political associations, I argue that attempts to cleanly distinguish civil from political associations can undermine democratic institutions and civil society. Additionally, the rise of modern surveillance technologies, developed largely by commercial entities, make state repression more efficient and dangerous than ever before.

To further show the risks of devaluing political associations for the sake of civil society, I examine the Second Red Scare, particularly Judith Shklar’s account of the United States Supreme Court case Dennis v. United States, 341 U.S. 494 (1951). Dennis upheld the convictions of members of the Communist Party of the U.S.A. for “conspiring to teach and advocate the violent overthrow of the government,” despite no evidence that the individuals were planning to participate in violence themselves (Belknap 2013, 1022).

While the decision in Dennis has since been de facto overruled, I revisit it for three reasons. First, the current presidential administration has led a litany of attacks on political expression. Examples include the denial of the Associated Press reporters to the press pool, investigating media companies and Universities for Diversity, Equity, and Inclusion practices, and detaining and deporting legal residents for their protected speech, including revoking the visas of people who posted critical comments of Charlie Kirk (Jurecic 2025). Most recently, anti-ICE activists, including one who was not even present at a noise demonstration where another participant shot an officer, were “tried and convicted under the Trump Department of Justice’s ‘counter-terrorism’ initiatives targeting “antifa” – short for antifascist – a decentralized movement the administration has officially categorized as a “domestic terrorist organization,” primarily because they produced, distributed, and transported zines (McMenamin et al. 2026; Goodman 2026). Dennis and the Second Red Scare, which echoed an earlier Red Scare that targeted anarchists, is particularly instructive for classical liberals, as they must reckon with the risk that fears of left-wing activism and redistributionist policies might be weaponized to undermine basic individual rights. Second, Dennis exemplifies the risks of embracing legal institutions as the means of escaping politics, rather than seeing legal institutions and liberal politics working in tandem in support of democracy and individual rights. Judith Shklar used the case to illustrate that, while liberals ought to champion legal norms and institutions, even liberal societies with well-established judiciary systems could engage in cases of political prosecution that were more common in totalitarian regimes. Given classical liberalism’s Cold War origins, it is particularly important for the classical liberal tradition to reckon with the history of anti-Communism and how fears of redistributionist policies have previously been used to justify state repression of individual rights.1 Third, the technological constraints on the federal government in the Second Red Scare, e.g., the rudimentary nature of wiretapping, serve as a helpful foil to the massive power of modern surveillance systems.

I return to the present moment by examining recent ways surveillance technologies, including artificial intelligence programs, have been used to undermine civil society and democracy.

While the First and Fourth Amendments inscribe in the American Constitution rights to free expression, association, and what we now consider privacy as protections against government surveillance and censorship, it is precisely when groups are seen as “political,” a slippery notion easily manipulated and selectively applied by public authorities, that those freedoms are most easily attacked. Recently, that surveillance has often been directed to Black Americans seeking redress as well as immigrants, echoing practices of the racializing slave surveillance practices distinctive of colonial America (Browne 2015). Additionally, the government and private industries are increasingly entangled as state surveillance and repression is only made possible by big data and artificial intelligence technologies developed and maintained by private companies, who, in turn, rely on government contracts (Bajwa 2026).

KEYWORDS: Democracy, Liberalism, Civil Society, Surveillance, Associations