Has accountability become a so-what question? On November 5, 2024, Donald Trump was elected to be President of the United States, again. In this article, we provide a different context for this historic election, by invoking the contemporary reference point for undemocracy: Hungary. Indeed, the UK daily The Guardian reminds us that Steve Bannon already called the Hungarian Prime Minister Viktor Orbán “Trump before Trump” (Smith, 2025). This is more than an analogy: evidence is growing by the day of the close links between Orban’s playbook and Trump’s, and the two allies openly admit to learning from each other. Analyzing how Orbán’s government was being held accountable in elections, what this accountability has yielded and how it evolved, helps us gain insight into the issues destabilizing even the largest democracies today.
Ahead of the 2022 election, the ruling Fidesz party led by Orbán was preparing for another win. The structure of the political landscape had been solidified since 2010 into a supermajority rule of increasingly extreme right Fidesz, headed by Viktor Orbán. With two-thirds majority in parliament, Fidesz changed the legal foundations of Hungarian democracy, including the Constitution, the judicial system and the regulatory landscape, whether in banking, welfare or education. Orbán consolidated the media and restricted the freedom of speech.
Attempts to mount a significant opposition, even for a minority government, had been unsuccessful, but Fidesz’s position weakened due to Covid and the Ukraine war. Sensing this opportunity, all opposition parties from left and even far right, formed a historical alliance and nominated candidates together, in all districts. By acting in unity, they believed they stood a good chance to finally unseat the government and stop the unraveling of democratic institutions.
Ordinary citizens and political actors all feared that a united opposition would not be enough if they cannot trust the election process. Reports by observers of the elections in 2014 and 2018 highlighted corruption through the governing party’s patronage networks (Mares & Young, 2019). Transport was organized to polling sites, and people were promised rewards if they voted Fidesz, or threatened. They had to show proof of their correct vote. One lesson for the upcoming 2022 elections was that local voting processes needed better monitoring.
Opposition parties realized there is a democratic institution they had not been using: the system of delegated election officials, often volunteers, who participate in the local election committees that oversee the voting process and count the votes on election day. In half of all polling sites, the non-governmental parties were not present. To address this blind spot, in just a few months, grassroots organizers mounted a massive campaign to recruit the missing 20,000 volunteers to serve throughout the country. There was special emphasis on remote areas, where events on the ground were more difficult to track and document, in the previous elections.
The effort was wildly successful. Over 20,000 people signed up, underwent training, were sworn in; and on election day, they made their way to small villages to serve on the committees — as monitors, poll workers and ballot counters all rolled into one. The election system in Hungary relies on a specific view of objectivity. Instead of prescribing an election committee that is politically “neutral”, it aims to create one that is “balanced”. Partisan poll workers participate from opposing parties, alongside municipal officials, themselves elected from different parties. The opposition parties were rebalancing these committees to monitor the local voting process.
Distrust but also discrediting of elemental democratic institutions such as elections is growing in liberal democracies of the Global North. Narratives of election fraud are no longer only associated with “weak” democracies; they have become a political strategy in all contexts. In the United States, voting results have been legally challenged in at least one presidential election since Donald Trump first came to power (Moynihan, 2022). A democratic voting procedure typically involves officials who manage the identification of voters, the casting of votes, counting the ballots, verifying the count, and inspecting a selection of counted votes. The US employs election officials, bipartisan election judges, and election auditors (Jacobs & Choate, 2022, p. 28). By attacking the results of the ballot count, the recount and its audit, Donald Trump has not only been delegitimizing the U.S. “administrative state” (Moynihan, 2022, p. 36) but also the institution of voting.
In the US, we see a far-right majority that positions itself against democracy and the “deep state”, distrust in the election is formulated as distrust of the state, not directly of a specific Democratic administration. In Hungary, however, distrust in the election spans the political spectrum, and directs itself at a state apparatus that is controlled by the Fidesz government — one that openly claims to have established an “illiberal democracy”.
As democracy is invoked and contested, the calls for accountability multiply. Citizens’ and parties’ renewed focus on election monitoring implies that if democracy is founded on systems of accountability, on checks and balances, then the means to reinforcing those democracies must surely have to do with accountability, too. As we discovered in our research that we describe here, however, accountability and democracy are not always reinforced through this monitoring tool. Rather, new relations of who is holding whom to account and for what, may emerge from its use. We gain firsthand insight into how democracy may or may not be able to manage the divisive differences in its members’ values: how it can sustain dissonance.
1 A Research Project on a Democratic Observation Technology
What does it mean to observe an election? Our research project investigates the citizens’ impulse to participate in monitoring and conducting elections. We approach this question from the perspective of technologies, putting aside theories of political science. In the spirit of Luhmannian discussions, which emphasize that society is composed of observers observing each other, we can consider election oversight as a technology of observation (Esposito & Stark, 2019). It provides a view into the running of democratic institutions, and into the very material of politics on the ground.
Unlike observation technologies that look at phenomena indirectly or through devices that access the “reality” (of the financial markets, for example), election monitoring is firsthand, real-time and immersive. In a twist to the distinction between first-order observation and second-order observation (of the first order), the election officials perform their second-order observation of the election process via first-order observation, by going to the site where the votes are cast and counting them — by being directly in the midst of the observed activity and observing it in its physical instance.
In a sense, it is the inverse of the observation technologies we tend to associate with the term, which try to grasp what is “going on” in a field through selected extracts from that field in scattered formats that are often asynchronous to what is happening, for example in financial markets. In Framing Finance (2009), Alex Preda emphasized how the reference to scientific practices of observation helped delimit finance as a separate realm that is only accessible to observers through technologies such as the stock ticker, which showed the “constant” flow of stock prices on streams of paper or screens. Credit ratings have been analyzed by Esposito and Stark (2019) as observation technologies of the second order. The election technology also has a format, but it tries to grasp those “goings on” by situating itself within them. Instead of getting the transcript, it participates in the conversation; instead of using a telescope, it goes to the phenomenon to size it up 1:1.
2 Volunteering to Witness and Perform Democracy
The volunteer election officials were driven by the desire to witness fraud and stop it. The spirit was, “This time, we are going to catch them”. The volunteers felt they needed to do what Elena Esposito has referred to as “first order observation”, from a Luhmannian perspective (Esposito, 2013). They wanted to see, what is really behind the poll numbers, the election results reported in the media? They wanted to witness the production of these numbers and catch how those could not be right. The technology of observation, the election official’s role, afforded this peeking behind.
In light of this mission, the election results were all the more bruising for the opposition and their volunteers: despite all efforts, the government won by a large margin, once again. Our question is: what happens when a technology of observation fails? When reality is observed “live”, people are seeing it “with their own eyes” — yet they do not see what they expected to see. What are the dynamics that are set in motion in this observational setting?
Having returned from their polling sites, the volunteers were invited to share their experience of surveying the election, on Facebook, in a designated, publicly visible group. Many volunteers posted online about what they saw. Soon after, the story blew up. It caught our eyes as researchers because a nationwide controversy erupted about the “accounts of vote counters”.
We understood the controversy when we started reviewing the posts one after the other. We must keep in mind that the volunteers’ main job was to count the votes. In Hungarian, they were called szavazatszámlálók — ballot counters. Indeed, their online posts started by sharing how this counting took place. “1070 people were registered in the electoral district and 800 people came to vote”. Then, it became much more. After saying how many people voted, the village population, and the results, they commonly stated that they “did not see any fraud”.
From these statements that belong in the world of verification, they make a surprising move: they seamlessly go on to give rich descriptions of the surroundings, the village, the people they encountered, sometimes in vignettes. They talk about the day spent in the committee, how they shared recipes, and enthusing about their encounters. But also:
There was no shortage of voters who could hardly move, could barely see, could barely sign [the electoral register], were rotting, or had basic hygiene problems.
From another post:
[These people] live in rickety, crumbling shacks at subsistence level […] there is poverty and ignorance.
Yet another post said:
As people from this tiny, tiny village wandered in, I had to wake up to the fact that most of them really are incredibly stupid.
Most detail came from those officials who took the ballot box to those citizens who were unable to come to the polling station. They talk about how when they walked in, there was an incredible stench, and the elderly lady was sitting with the TV on, blasting the national channel of government news, and dirt everywhere… or, that the little village was very tidy, with flowers everywhere.
These descriptions go well beyond giving an account of the official, mostly quantitative task. They take a qualitative turn that fits the call to share their “experience”. The implication, often pronounced, was: “no wonder” “these people” voted for Orbán.
These citizens, fueled by the virtue of upholding democracy, were primed to see fraud. But when they went there, they did not see fraud. And still Fidesz won. Over and over, when they talk about their experience on social media or in our research interviews, they are trying to understand this paradoxical situation. Most were not prepared for it, and experienced huge disappointment after such an effort on their part. The training consisted of 8 modules that had to be passed online; then they took two self-funded trips to their assigned polling site hundreds of kilometers away. They often add, “Fidesz did not need even to cheat in order to win. The election was lost elsewhere”. Or that “the fraud had been committed elsewhere” in the political system.
2.1 Accessing First-order Reality by Observing It
Can we say that this first-order reality of the election was not only observed, but that it was also created by these volunteer observers? In the strict sense of overseeing and securing the vote count, they were there to see this reality. But since volunteers also participated in the voting process by sitting in the committee, their observation looks more like the creation of that reality. Simply put, the election officials themselves made the election happen by handing out the ballots, checking signatures, or managing the waiting lines. Even the act of counting the votes is part of making this first-order reality, and not only checking its reality. If we follow a close Latourian approach of tracing how we move from people casting their ballots on election day to the “second-order reality” of election results aggregated into numbers and graphs and onto screens worldwide, the ballot counters are hardly the fabricators of the second order. At the same time, they were doing transformations to the paper ballots, sorting them, categorizing them (valid-invalid, by party) and arranging them in different piles, associating certain sheets with certain other sheets, counting them, recording the announced numbers in writing, recounting, signing documents. In this second sense, too, the volunteers were creating the first-order reality of the validated election.
The election observations verified the counting and the process, and they lead to accounting on the part of the observers. First, we have suggested elsewhere (Györke & Vargha, 2022) that the vote count and monitoring can be seen as an audit, as a ritual of verification: going through the motions to check that a process has taken place in conformity with the rules, and that the numbers so produced are correct. Indeed, the notion of living in an “audit society” (Power, 1997) captures the sentiment that the practice of auditing has spread outside business into most areas of life, as a new form of power. Similarly to financial regulators being second-order observers of finance as a domain, of traders’ activities, election auditors can be seen as second-order observers of their domain of politics, of voting activity.
Second, the ballot counters produced accountings and accounts as the output of their audit-like job. One set of accounts was done in their official role: they had to report a set of numbers, how many votes were cast, how many invalid, sign off on these numbers and on the official minutes. But they produced other types of accounts as well, such as the online posts, which spilled over and above the framing of their mainly quantitative task. As we saw above, these were qualitatively different accounts, and resided in different formats and different social and physical spaces from each other.
2.2 A Participant Observation Technology
The reason these in-depth accounts emerged is partly that citizen volunteers acted as participant observers. As we detailed above, they not only observed the voting process, like those Trump supporters recruited by the Republican party in the US, but they were effectively working in the election. They were immersed in the setting, observing what they were officially asked to see, and everything else. They were ethnographers.
We want to highlight three aspects of being human participants of the observation technology. First, democratic accountability shifted as a result. We argue that through their audit task of counting votes and verifying the election process, and through the accounts they produced, these citizens created new accountability relations. Originally, they engaged in this civil society effort in order to hold their government accountable. But in the end, as they had not been able to witness the fraud they expected, they held society accountable. They did that by telling their stories about all the things they saw, beyond the official, beyond the quantitative. What those voters are like in the other parts of the country, and how they are so different from us; and by suggesting implicitly or explicitly, that this difference is why those people voted for the government. The accountability relations we initially assumed had shifted.
In holding society accountable, these participant observers mobilized tacit theories explaining this unexpected result, of an illiberal, undemocratic regime winning reelection without voting fraud. Instead of being fully-fledged, academic, social theories, these are vernacular theories that ordinary people implicitly generate when they give account to each other. Instead of holding them in stock as coherent wholes, such theories materialize when the moment calls for them, in the interpretive act of remedying socially breaking situations (Garfinkel, 1967).
The online posts by participant observers elaborate or indicate various theories about voters’ behavior, and about how state power operates. As voters, they would describe Hungarians as “submissive” or “ignorant folk”. They did not always blame the people. For example, when vote counters brought up that the television was running in the background with the government channel on, they mentioned media control — the notion that by dismantling independent media systematically and over time, Fidesz succeeded in creating a national media bubble in which only the government messaging gets across. Others talked about clientelism networks, handouts and town beautification projects for mayors loyal to the government.
The second aspect of participant observation is that many of the accounts that citizen auditors produced are truly and classically ethnographic: in a sense, they are written from a somewhat colonialist perspective. What emerged is that typically urban, educated professionals and students were describing “The Other”, the country folk. Witnessing how the country lives, and reporting back about all the things they had seen.
Election officials who posted on social media were not locals (with a few exceptions), and they were identifiably of the political opposition. This instantly made them aliens, arriving into a close-knit community of tiny polling districts. A number of factors may have contributed further to the social contrast between observer and observed. These could have amplified the “othering” dynamic. For instance, the democratic challenge was to fill the election committees primarily in small settlements, where the opposition is strongly under-represented. This meant that volunteers were bound to come from larger regional centers or the capital. The online-heavy recruitment and training of volunteers, and the time and resources required to travel and lodge at distant polling sites, could all have contributed to the geographic and socio-economic distance of the volunteers from locals.
Being the odd ones out, the volunteers were motivated to fit in for the day. Like most ethnographers, they made efforts to integrate, knowing that their stay is temporary. It was clear from our interviews that they did not speak the same language and felt culturally distant from the locals. However, they managed to find common ground in their conversation topics, and they could also rely on the official language of their election job.
The third aspect is that observation eventually became circular, because the participant observers were also being observed by others — due to the public nature of social media accounts. The circularity of observation, as Esposito (2013) and Stark (2013) have emphasized is key to understanding society as observation relations. In this case, it emerged from the shifting of democratic accountability, which led to the colonialist “othering” descriptions. After an accumulation of social media posts lamenting the poor conditions and ignorance in the “countryside”, an outcry followed from people who protested to “being described” in this way. These users posted vitriolic responses disputing the deprecating descriptions, quickly supported by the national, government-backed media. They called out the authors of the original posts for being condescending and ignorant themselves. This way, the self-appointed adjudicators of Hungarian society were themselves held to account.
3 Democracy: Organizing Diversity, Managing Dissonance
We bring two provocations from this setting. One, what does the vote counting exercise say about the state of democracy? Lawfulness and democracy become weaponized by any and all parties. In the Hungarian case, we saw one group desperately clinging to democratic institutions and use them in a last-ditched effort to uproot the undemocratic. In the US case, for instance in the Republican National Committee’s drive to recruit election observers, we see a cynical upholding of democratic institutions by political groups who otherwise dismiss them. There was one thing in common, however. The institution of the election is distrusted to the point that, there was a sense ahead of the vote that the voting cannot be right unless one’s own party wins. Indeed, in the US, election officials are now harassed on and off the job by Republican supporters, and effectively held responsible for any unfavorable election outcomes (Wong, 2024).
How can we rebuild trust in these democratic institutions? Later in 2022, Members of the European Parliament (MEPs) adopted a report which calls Hungary a “hybrid regime of electoral autocracy”, i.e. a constitutional system in which elections occur, but respect for democratic norms and standards is absent (European Parliament, 2022). Is accountability still the right entry point? Will more monitoring and verification help reverse what the report identifies as “democratic backsliding”?
The accountability exercise undertaken by citizens in Hungary ultimately reinforced their distrust. They became disillusioned with the notion of fair elections. For many if not most, their accountability work only helped legitimate an autocratic regime. More precisely, their conclusions suggest that beyond election day procedures, the notion of fair elections must imperatively include a scrutiny of the campaign phase and of access to media. Ballot counters’ accounts of their interactions with voters repeatedly mentioned citizens deprived of alternatives, several steps prior to casting their votes. In addition, the publicness of social media inevitably generated a circular accountability, where those who speak about their job of assuring the election are themselves scrutinized.
No amount of citizen oversight will make the election trusted if their fairness is seen to be a marginal issue or a laundering device behind structural inequalities of representation. This is the limit of oversight here. As Esposito (2013) explains, in Luhmann’s view of social systems, second-order observation — observing the observers, monitoring the elections — is important as it creates reflexivity. This is how the systems adapt, by regulating and by reacting to regulation, for instance. Here we see that second order observation becomes discredited, distrusted, trivialized.
If elections and their monitoring are no longer the key moments of democracy but the mere end of an unfair pipeline, the question becomes: how can we rebuild the democratic institutions themselves?
The second provocation then, is how to address a profound sense of political dissonance, for new democratic institutions. How can we navigate the day-to-day situations of working together, face-to-face with deep political divides? How does democracy handle profound value differences in the everyday — and looking forward, what does it take to handle them?
We reach to the notion of organizational technologies here. Fields such as organization studies and economic sociology have paid attention to the stabilized, durable configurations of people, procedures, objects and ideas that can together be called technologies. These technologies participate in making organized action possible, whether it means trading in markets (Muniesa et al., 2007) controlling a company (Yates, 1993), or enacting democratic politics in parliament architectures (Dányi, 2017).
We consider that democracy is itself an organizational technology, one that is meant to organize diversity.1 The diversity of interests and opinions, of ways of being and thinking. The democratic institutions are engineered to prop up a certain vision of society that maintains diversity but operates with methods for containing it. What is often missing is the street-level view of how democratic societies manage that diversity. This is becoming a vital issue in contemporary settings where diversity is amplified and experienced as profound dissonance.
In the sociology of organizations, studies have explored how “dissonance” is managed within organizations (Stark, 2009; Antal et al., 2015). When some in a startup pressure for financial success while others insist on taking the time to design software properly, will one approach win out? Instead of facing conflicting professional values head-on, research has found that organizations have the capacity to find ways to maintain those differences and work them into new solutions.
What the Hungarian opposition volunteers experienced on election day, working in their committees, was a problem of managing dissonance in a democracy. They had to work with election officials sent by the government party, feared as all-powerful and oppressive, and with the elected local municipal officials, often members of the extreme-right governing party.
How did these election volunteers last through the day, surrounded by those whose values and beliefs they profoundly disagreed with, and vice versa? And how did they deal with the fact that their discord was public knowledge? The clue to the answer is that election monitoring and counting votes are situated activities, so that volunteers were immersed in the to do’s of the day, and in the flow of one situation into the next. Besides this situated inertia, they also wanted to get along for the time spent in close quarters. And they did. The accounts on social media and in interviews pointed out how well they could work with the committee. They found that the other vote monitor-election officials from the government party turned out to be mostly nice people, with whom they could discover common interests (kids and recipes) and solve task-related issues. Some interviewees remarked that they found themselves joking with some locals on the committee throughout the whole day.
My counting partner was a self-confirmed Fidesz fan. He was a terribly nice guy. He was smart, we talked a lot, we teased each other, he was the most positive character in the whole story.
This looks to be a different way of managing dissonance than what organizational research has highlighted. Finding common ground in areas that are arguably inconsequential, while avoiding the well-known controversial topics (democracy, war, immigration) as happened here, is different from reaching a new working order where the dissonant views are relatively fulfilled. Ground-level democracy has a lot of tasks: to hold power accountable, to get along, and to create new solutions of habiting together.
Here is our provocation: if we are to live with profound dissonance reverberating in polarized communities, what are better ways of managing it within democratic principles? How can we move forward and co-exist? Or is this how we lose democracy, the institution: through small talk, glossing over the fact that the hands that baked the same cake recipe may cast their votes for radically different versions of society, not all of them democratic? Viktor Orbán proudly calls Hungary “a petri dish for illiberalism” (Smith, 2025). Can it also become the petri dish for renewing democratic institutions for an age of dissonance?
References
Antal, A.B., Hutter, M., & Stark, D. (Eds.). (2015). Moments of Valuation: Exploring Sites of Dissonance. Oxford: Oxford University Press.
Dányi, E. (2017). Politics beyond Words: Ethnography of Political Institutions. In R. Wodak & B. Forchtner (Eds.), The Routledge Handbook of Language and Politics (pp. 291–306). London: Routledge. https://doi.org/10.4324/9781315183718-22
Esposito, E. (2013). Economic Circularities and Second-Order Observation: The Reality of Ratings. Sociologica, 7(2), 1–10. https://doi.org10.2383/74851
Esposito, E., & Stark, D. (2019). What’s Observed in a Rating? Rankings as Orientation in the Face of Uncertainty. Theory, Culture and Society, 36(4), 3–26. https://doi.org/10.1177/0263276419826276
Garfinkel, H. (1967). Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Györke, M., & Vargha, Z. (2022). From Counting to Giving Account: Citizens Auditing the 2022 Hungarian Parliamentary Elections. ESCP Impact Papers, 3, 188–197.
Mares, I., & Young, L. (2019). Varieties of Clientelism in Hungarian Elections. Comparative Politics, 51(3), 449–471. https://doi.org/10.5129/001041519X15647434969911
Moynihan, D. (2022), Delegitimization, Deconstruction and Control: Undermining the Administrative State. The ANNALS of the American Academy of Political and Social Science, 699(1), 36–49. https://doi.org/10.1177/00027162211069723
Muniesa, F., Millo, Y., & Callon, M. (2007). An Introduction to Market Devices. The Sociological Review, 55(2_suppl), 1–12. https://doi.org/10.1111/j.1467-954X.2007.00727.x
Power, M. (1997), The Audit Society: Rituals of Verification. Oxford: Oxford University Press.
Preda, A. (2009). Framing Finance: The Boundaries of Markets and Modern Capitalism. Chicago: University of Chicago Press.
Smith, D. (2025). “In a real sense, US democracy has died”: How Trump is Emulating Hungary’s Orbán. The Guardian, February 7. https://www.theguardian.com/us-news/2025/feb/07/trump-viktor-orban-electoral-autocracy
European Parliament (2022). MEPs: Hungary Can No Longer Be Considered a Full Democracy. Press Release, September 15. https://www.europarl.europa.eu/news/en/press-room/20220909IPR40137/meps-hungary-can-no-longer-be-considered-a-full-democracy
Stark, D. (2009). The Sense of Dissonance: Accounts of Worth in Economic Life. Princeton, NJ: Princeton University Press.
Stark, D. (2013). Observing Finance as a Network of Observations. Sociologica, 7(2), 21–25. https://doi.org/10.2383/74854
Wong, M. (2024). “Stop counting votes, or we’re going to murder your children”. Election Officials are Under Siege. The Atlantic, October 25. https://www.theatlantic.com/technology/archive/2024/10/election-workers-threats-trump/680362/
Yates, J. (1993). Control Through Communication: The Rise of System in American Management. Baltimore: Johns Hopkins University Press.
Thanks to László Bruszt for emphasizing this formulation in discussing our argument.↩︎