Cyborg political machines

Colombian professional political consultants couple information technologies and local political brokering to circumvent strict voter privacy regulations that limit campaigns’ access to voters’ personal data. I argue that political consultants use information technologies to bolster traditional vertical, personality-centered political organizations, and to produce tightly controlled “cyborg political machines.” I challenge widespread notions that oppose media-based politics to traditional face-to-face politics (known also as clientelism). Instead, I show that although political elites introduced American political marketing methods hoping to modernize campaigns, the American way provided a new framework to preserve traditional authoritarian political arrangements after the extensive democratic reforms of the early 1990s.

In a private room in a British-style gentlemen's club in Bogotá, the owners of a Colombian political strategy firm, which I will call Pragmatica, 1 pitched their services to a candidate running for Congress. These outfits are staffed with public relations experts and lawyers who plan campaigns on behalf of their clients, political candidates. The owners of the firm-a constitutional lawyer and self-proclaimed "persuasion expert, " and a lawyer with a master's degree in security studies-sold to the candidate a standardized campaign management package comprised of seven key variables. The seven variables included what the consultants termed "image, message, issue, political structure, political chess, time management, and money. " One of the consultants explained to the candidate what the variables meant: The first three variables will help us choose the main issue that voters will know you for-they will help us come up with a straightforward campaign message, and design a media strategy to appeal to voters easily influenced by media, mostly middle-and upper-class voters. The last four variables will help us organize the resources of your campaign, including leveraging the political friendships that you currently possess. Your political friends can help us access their votes in low-income neighborhoods.
The political consultants concluded their pitch by emphasizing that their campaign management methodology was highly effective. I argue that the standard "magic formulas" that Colombian professional political consultants sell to campaigns (such as the seven variables that guarantee success) use information technologies to bolster local vertical political networks, instead of replacing them. Despite many political reforms intended to weaken the power of party bosses and their brokers (commonly known as political machines), and to bring representatives and citizens in closer contact through new media technologies, electoral politics remains tied up with traditional local powers (see Arjona 2016;López 2016). Although much of the scholarship on Colombia has mostly focused on the violent mechanisms these local groups exert, I contribute a different angle: I question the role of political marketing in boosting these local political networks.
Pragmatica's pitch to the candidate encapsulates an emergent information-driven approach to selling politics to voters that has been imported from the United States, and that has been adapted to fit Colombia's protective information regulations. These regulations include Habeas Data, 2 a law that prohibits campaigns from accessing voters' personal data without their explicit consent, and electoral laws that limit campaigns' media exposure. Colombian political consultants rely on local political brokers to legally collect voters' personal information and reach out to them. Consultants also use publicly available voting rolls, demographic records, and legal tools to supervise-or "discipline"-the work of these brokers.
Inspired by Donna Haraway's "Cyborg Manifesto" (1990), which contests biological essentialism and Cartesian dualism, I use the term "cyborg political machines" to underscore that present-day face-to-face politics and media-based politics are indistinguishable. Recent analyses of Latin American political communication argue that electoral campaigns have undergone a process of "hybridization" resulting from the introduction of American campaigning to the region. Carlos de la Torre and Catherine Conaghan (2009), for example, deploy the concept of the "hybrid campaign" to explain the persistence of populism, authoritarianism, and clientelistic politics (what they see as "old political technologies") despite the rise of mediabased politics and telegenic candidates (what they in contrast represent as "new political technologies"). The authors conclude that "each country's political culture and past shape the ways in which candidates and their campaigns frame their character appeals to the electorate" (ibid.: 348).
I deploy the concept of the cyborg political machine to contest the idea that Latin American political cultures are in essence populist and authoritarian, and that media-based political technologies originating in the global Northwest, by contrast, are not. Through the ethnographic study of Colombian political consulting, I find instead that Colombian elites could easily translate political marketing methods to bolster political machines because both marketing and machine politics rely on similar top-down information practices that aim to persuade and influence voters, not rationally debate with them.
Recent works analyzing electoral campaigns in the United States argue that data-intensive, media-based political campaigning relies on top-down information management practices to bolster participation on the ground, but in ways that align with the strategic directives of electing a candidate (Nielsen 2012;Stromer-Galley 2014). Data-intensive campaigning depends on consolidating large information sets, including demographics, and citizens' voting and consumption histories, to predict voters' choices and persuade them to vote. Daniel Kreiss (2016) further argues that the political effectiveness of these voter information systems depends upon creating social and political practices to keep information updated.
The case of Colombian political consulting shows that in contexts where voter information is not readily available, old-fashioned political brokering is indispensable to political marketing. This observation suggests that American-inspired political technologies may be also acting to reinforce what scholars have termed illiberal democracy-systems in which political elites preserve democratic form, but discourage public debate and the exercise of civil rights (Arias and Goldstein 2010; see also Holston 2008).
I support my argument with ethnographic evidence collected since 2007, when I started following then President Uribe's effective political communication strategy. Despite the many scandals that marred his presidency, nothing seemed to tarnish his high public approval ratings. Puzzled by this phenomenon, I started digging deeper into the flourishing industry of Colombian political consulting.
I went back to Bogotá in time for the national election of 2010, where I spent twelve months working alongside different kinds of electoral experts, including electoral observers, journalists covering the election, and political consultants of different types. I worked as the coordinator for the International Electoral Observation for a local NGO, Misión de Observación Electoral (MOE), where I learned about electoral regulation, and I gained access to regulatory state entities, international democracy-building organizations, and to political parties. I also shadowed the work of Pragmatica, and through them I accessed a convoluted network of professionals that included pollsters, advertisers, digital marketers, campaign managers, press agents, politicians, journalists, hackers, political operatives, and local political brokers. I also interviewed the American political consultants who first exported the industry to Colombia in the 1980s.

Cyborg political machines
Professional political consulting entails many forms of expertise, ranging from general strategic consulting to media production. Pragmatica is a generalist political strategy firm that advises the overarching strategy of campaigns across parties, and that convenes various experts to meet campaigns' needs. Pragmatica's consultants differentiate themselves from advertising companies that turn into political consulting firms during election season. Pragmatica's services go beyond producing political advertisement-they combine strategic communication with legal and lobbying services.
I first met Pragmatica's founding partners and lead strategists in 2009 in Bogotá, in a conversation that started off in their office, and ended in a restaurant. At the time, Pragmatica was advising a primary presidential campaign, and several campaigns for Congress from candidates from different regions and different parties (except the leftist party Polo Democrático and the Green Party, which the consultants considered to fall outside their wide ideological spectrum).
Many meals together followed. Much of the work I did with Pragmatica involved driving around Bogotá with the lead consultants, and sitting in meetings with candidates and with other election professionals that Pragmatica convened to work in different campaigns. In these meetings, the consultants would repeatedly emphasize that Colombian electoral regulations strictly constrain campaigning timelines and budgets. The consultants would argue that winning elections required an expert team working in coordination to leverage the support of powerful regional political bosses (padrinos políticos), use existing regulations to a campaign's advantage, and design media products in touch with target voters. Although at first I was concerned that the many meetings were taking away time from the real action of campaigns, in these meetings cyborg political machines took shape.
The concept of the cyborg political machine helps explain enduring political informality in the wake of vast institutional transformations and rapidly changing media infrastructures. Cyborg political machines resemble what scholars studying campaigns in the United States have described as "campaign assemblages": networks of actors, organizations, and technologies that come together for the period of a campaign to work in concerted action to elect a candidate (Nielsen 2012;Stromer-Galley 2014). However, unlike the campaign networks of the United States, cyborg political machines not only work in concerted action to elect a candidate, but they also combine new information technologies and traditional political forms that work in concerted action to exercise vertical information "disciplines, " and preserve hierarchical political networks.
Anthropologists working in very different contexts have explained the pervasiveness of vertical patron-client politics despite profound democratic reforms by questioning the universality of the concept of democracy (Nugent 2008). These works generally have argued against understanding all democracies in terms of Northwestern historical trajectories (Witsoe 2013), and also argue against the idea that clientelism is an anomalous, asymmetrical, socially binding form of social exchange (Roniger and Günes-Ayata 1994) arising from defective or incipient market economies (Gellner and Waterbury 1977).
Specifically, ethnographic accounts have challenged a dominant economic understanding of voters as calculating individuals seeking maximum benefit (see Roudakova 2008), to show that social identities, like religion and gender, also play an important part in how voters cast (or abstain from casting) their ballot (Gutmann 2002;Banerjee 2011). Ethnographic works on clientelism have shown, for example, the mechanisms by which informal political networks transform everyday social relations into political authority (Auyero 2000;Ansell 2014); the "moral idioms" and normative political imaginations shaped in these networks (Piliavsky 2014); and the circulation of favors and goods as socially binding practices that define people's experience of the state (Ledeneva 1998).
Like these ethnographic works, I question the persistence of clientelism despite the rise of marketing-based campaigning. Similarly to the clientelism/democracy opposition, works in political marketing have generally represented clientelism as an enduring, premodern political practice that threatens open, horizontal, information-driven forms of political competition (Hallin and Papathanassopoulos 2002;Pasotti 2010). Media, the argument goes, are a democratizing force that provides voters with a platform to communicate directly with their representatives, therefore weakening the shady operations of local political brokers, and favorbased politics. However, when individual agendas and self-interests co-opt these information systems, scholars warn, media reinforce vertical political relations.
The case of Colombian professional political consulting suggests that political marketing and traditional Colombian political networks are similarly hierarchical, and therefore work well together. Both are closely supervised pyramids that instrumentalize voters' support to advance their own goals of electing a candidate, and that preclude voters' input in a campaign's decision-making process (Leal Buitrago and Dávila López. 1990).
Pragmatica belongs to an emerging industry of Colombian political consulting molded to the image of the American public relations industry. Similarly to their American counterparts, Colombian political consulting firms routinely face the challenge of collecting detailed voter information and translating these information insights into effective persuasion campaigns. Insiders to the political marketing industry observe that nobody knows with precision what works to bring voters to the polls. Political consultants, both in the United States and in Colombia, sell to candidates the belief that "their tricks somehow bring order to the chaos of a campaign" (Sabato 1981: 11). Here, order is a key word. Although these "tricks" may not be effective in persuading voters, the methodologies that political consultants sell to campaigns, like Pragmatica's "seven variables" formula, contribute to solidify political practices that privilege top-down management over horizontal deliberation.
With this observation, however, I am not suggesting that the adoption of media technologies results in the erosion of local democracies and horizontal decision making. Normative scholarly perspectives often posit true political consensus as the result of meaningful horizontal debate among stakeholders (Habermas 1989) and "genuine collective action" (Bourdieu 2004) among individuals who can equally influence politics. These works oppose "true" participation to mediatizationcommonly understood as the technification of the political process.
Works in mediatization often emphasize its outcomes-like when it results in the oversimplification of complex political processes into televisual political slogans; or when it interpellates citizens as passive spectators of entertainment politics (Hall, Goldstein, and Ingram 2016), in which hyperbolic candidates grotesquely incarnate the tastes, lifestyles, and social realities of voters instead of representing their interests (Rao 2010;Lempert and Silverstein 2012).
In place of questioning mediatization as an outcome, I follow other anthropologists who understand it as a political technology modulated by and productive of local politics (Paley 2004;Juris 2012;Graan 2016). I find that Colombian branded, hyperbolic candidates do not live in the media sphere alone, but that the information practices that go into making these celebrity-style candidates create political conditions that renew the authority of on-the-ground political networks.
The literature on populism is useful to explain the connection between mediatization, marketing, and clientelism. As Ernesto Laclau (2005) argues, populism is a "logic of articulation" that clusters together the experience of disenfranchisement of different social groups with very different types of political, economic, and social grievances (or claims). Populist discourse strips away these claims from their social context, and reduces them to chauvinistic slogans. Laclau further explains that populist discourse aims to change citizens' semiotic relation to democracy by redefining democratic concepts (like equality, fairness, legitimacy, etc.) in terms of the experiences of disenfranchisement of the followers of the populist leader. Adding to Laclau's point, Jeffrey Witsoe's (2013) work in Bihar, India, underscores that cultural practices and the distribution of resources support populist discourse. His study highlights that political loyalties based on caste and religious affiliation both inflect populist rhetoric, and organize the distribution of economic and political resources. Witsoe concludes that this marriage between discourse and practice animates a populist bureaucracy that shapes citizens' experience with democracy and the state.
I see a few differences and many similarities between populism and political marketing. Unlike populism, political marketing is a neoliberal technology that celebrates the individual and appeals to voters' lifestyles, life experiences, and social values instead of appealing directly to their unfulfilled social claims. 3 However, political marketing also relies on a similar "logic of articulation" to populism.
Political marketing is a communication technology that aims to sell politics to as many buyers as are needed to win an election. Marketers style the image of candidates and campaign messages to persuade different segments of voters using familiar tropes that resonate with voters' cultural contexts and symbolic worlds (Applbaum 2004). Experts in branding also produce oversimplified political messages designed to connect emotionally with different segments of voters (Lempert and Silverstein 2012), and to create powerful, easily understandable, relatable storylines. Similar to populism, these catch-all political stories often encompass oversimplified political identities that presume an abstract opponent (Brader 2006)-a 3. I refrain from using the term "neopopulism" coined by scholars of Latin America to describe the intersection of populism and media-based politics (Barczak 2001;Waisbord 2003;Weyland 2003;Cristina de la Torre 2005). This body of literature generally describes neopopulists as political leaders who deploy public opinion and marketing to legitimize their actions-instead of effectively including citizens in government-and who differentiate from their populist predecessors in their aversion to state welfarism and party structures. The Colombian case shows that the media effectiveness of neopopulist leaders also depends on bolstering local political networks inherited from traditional party politics that work in concerted action to amplify their messages, and that materialize those messages in political practice.
threat that the candidate stands against (think of Obama's message of "Change" against the political establishment; Trump's "Make America Great Again, " which vilified immigrants and Muslims; or Brexit's "Take Back Control" directed against the European Union). Like populism, political marketing is also mostly effective when coupled with on-the-ground political practices and political networks. Political consultants craft oversimplified campaign messages that can be easily adapted to convince different segments of voters. Usually, campaigns rely on political networks and trusted leaders in a community to adapt and credibly deliver a campaign's messages. As I will show with the case of Uribe in the last part of this essay, in return, these local political networks gain a platform, a microphone to voice their own views, and they access resources once a candidate is in office.

The birth of modern political campaigning in Colombia
The Colombian Constitution of 1991 transformed the country's political layout. As in many other countries in Latin America, the World Bank and the International Monetary Fund prescribed a set of neoliberal reforms to strengthen the Colombian economic and political systems, and to ultimately (hopefully) bring peace to a country marred by the War on Drugs and a guerrilla war dating back to the 1960s. Those reforms included the opening of the Colombian market to foreign investment, and the dissolution of the traditional two-party system. To promote more direct civic participation in government, which reformers believed would end ongoing political violence by including traditionally excluded groups, the newly drafted Constitution of 1991 guaranteed freedom of association and the possibility of forming new political parties. As in many other countries, international organizations introduced the notion that electoral reforms could remediate a "broken" society and prepare citizens for democracy (Coles 2007).
This "new" system paradoxically formalized an existing tradition of strongmen and pyramidal, candidate-centered political structures within the traditional parties, what Eduardo Pizarro Leon-Gómez (2002) calls "empresas electorales" (electoral businesses). The reforms freed candidates, party brokers, and voters from the two parties, which had traditionally controlled political resources and access to voters. Candidates now free to leave their parties founded their own parties, and by 2001, sixty-seven parties and political movements were officially registered (Redacción, El Tiempo 2001). Since 1991, governments have passed many laws to regulate this messy party system while also preserving the participatory spirit that informed the reforms in the first place (in 1994, 2003, 2005, 2009, 2011, 2015). These modifications aimed to curtail "electoral businesses" by forcing parties and campaigns to publicly account for their finances, by establishing rules for party membership and voting thresholds that parties have to meet in order to keep their legal status, and by limiting the time, places, and methods for electoral campaigning to guarantee fair media access.
Parallel to this institutional shift, electoral campaigning also changed. American consultants introduced "modern political campaigning": an information-based form of campaigning that focuses on public relations, on marketing a candidate's image and message. Starting in the mid-1980s, Colombian governments hired American experts in communications and crisis management to mediate the strained relationship between Washington, DC and Bogotá resulting from Colombia's cocaine boom. 4 Colombian governments depended on international credit, especially American assistance, yet the US Congress was reluctant to commit to more aid once the drug epidemic hit American streets. President Virgilio Barco (1986-90) hired Sawyer Miller to work on his presidential campaign and, once in office, extended their contract to manage Colombia's reputation in the United States.
I interviewed Jack Leslie, current chairman of Weber Shandwick, one of the largest public relation firms in the world, and who was then president of Sawyer Miller. Virgilio Barco met Leslie on a trip to Manhattan in 1985. Leslie explained that Barco hired an expensive international team to manage his presidential campaign, believing that outsiders would "modernize" Colombian campaigning. Barco and Leslie shared the belief that television-centered politics would transform a traditionally exclusionary political culture. They believed that public relations and media management could replace political brokers, and take away political power from the hands of the "mafiosos" who had infiltrated traditional clientelistic political networks, a phenomenon termed "narcoclientelism" (Restrepo 2006).
Barco's campaign successfully integrated public relations and party politics. Later in our interview, Leslie remembered the work of Eduardo Mestre, who coordinated the political strategy with existing political networks in the regions, and whom Barco later demoted because of Mestre's ties with the Cali Cartel. In our interview, Leslie repeatedly used the term "modern" to describe media-based campaigning as opposed to party politics, as if to distance himself from Mestre and his work in the regions. However, Leslie's usage of the term "modern" can be misleading-it conceals the fact that American public relations experts relied on local political networks to spread campaign messages and persuade voters. I do not think, however, that Leslie intended to misrepresent his work. Instead, the term "modern" carries a heavy moral baggage that obfuscates new technologies' interactions with longstanding political practices.
"Modern political campaigning" was a term born in the United States with the rise of the public relations industry in the 1920s, and the decline of political machines (Schudson 1998). "Modern political campaigning" understands campaigning as a professional, rational practice aimed at managing individuals' preferences, and promoting the direct relationship between candidates and individual, anonymous citizens through media. The term "modern" exists in opposition to notions of a "premodern" form of campaigning governed by face-to-face relations, political hierarchies, and loyalty ties.
When the Americans arrived in Colombia, they introduced "modern" political campaigning technologies, with their moral underpinnings. Since Barco, presidential campaigns in Colombia have usually hired American consultants who bring their own methodologies, and who train Colombian political professionals. The high-profile Colombian political consultants whom I interviewed could easily trace their lineage back to a specific American consultant-often presenting these connections as credentials for their authority. This informal training process has shaped an increasingly professionalizing industry of Colombian political consulting that has adapted these "modern" methodologies and moral imaginaries to the Colombian context. Jaime Bermúdez, who advised the communication strategy for Álvaro Uribe's presidential campaigns in 2002 and 2006, and for his government, described this adaptation process to me in an interview. Bermúdez humorously used the concept of "platanización" ("bananification") to highlight the challenges of adapting "modern" methods to the context of Colombia's tropical, banana-republic style politics: What do the gringos know that we don't know? Or how to platanizar what the gringo team is proposing to us? The methods imported are very simple, they are just methods to generate information. After, with discipline and shrewdness, you can translate those methods into an important [communication] strategy. If you do not have that feeling or that intuition, which usually comes from the candidate, to platanizar, to adopt, to translate, it is very difficult for the strategy to work well.
Colombian political consultants widely use two "vernacular" categories to describe voters that exemplify this process of translation. Americans use the concepts of "soft support" (or swing vote) and "hard votes" (partisan base) to calculate voters' commitment to a candidate. The Colombian equivalents are the "opinion vote" and the "structure or bounded vote" (voto de estructura or voto amarrado). Unlike the American concepts, not only does the opinion vote/structure vote divide refer to the strength of a voter's commitment to a candidate, but political consultants use the two categories to describe voters' behaviors in terms of historical configurations of class and region.

The opinion vote/structure vote
In the presidential election of 2010, Juan Manuel Santos, considered at the time President Álvaro Uribe's political heir, ran against Antanas Mockus, a scholarturned-politician, and former mayor of Bogotá. For the first time, campaigns relied on social media, especially Facebook and Twitter. Mainly using social media to campaign, Mockus galvanized a base of young supporters and liberals from the largest cities with an anticorruption, antiestablishment message; what became known as la ola verde ("the green wave"). Santos campaigned exploiting the existing support for Uribe to his advantage. In the days leading to the first round of the presidential election, the enthusiasm for Mockus in social media overshadowed that for Santos. To many, 2010 would be an inflection point in Colombian politics, a move away from Uribe's war-intensive, corrupt presidency to the construction of a political culture respectful of the rule of law heralded by Mockus. In a surprising landslide win (surprising at least to social media users), Santos won the presidency. Left in shock, Mockus' supporters and sympathizers viewed Santos' win to be symptomatic of an enduring premodern, feudal, political culture ensconced in the regions that privileges favors, self-interest, and friendships over collective well-being.
Political insiders, however, did not find Santos' win surprising. In the numerous conversations and meetings I had with campaigners and election professionals of all ideological stripes, they explained it by conceptually dividing the Colombian electorate into the "opinion vote" and the "structure vote. " Election experts explained that Santos won because he ran a "disciplined" campaign that had "structure votes" on his side: voters who on election day reliably turned out to vote because their vote guarantees benefits. By contrast, these experts explained, "opinion votes" make noise on media but do not reliably turn out to vote because they do not have any stakes invested in the election.
Generally, a campaign's "discipline" means its ability to control information. In Colombia, the concept of electoral "discipline" evokes class ideology. Election professionals consider the undisciplined "opinion vote" to be a middle-and upperclass, educated, urban voter who casts her vote following her consciousness, and who can be easily swayed through arguments and emotionally charged, mediabased persuasion techniques. In contrast, election professionals consider the rural and urban poor disciplined "structure voters, " who "vote with their stomach, " as a campaign manager put it, and whose loyalty is up for bid.
Regional histories and social stereotypes inform the assumptions at work in the opinion/structure vote divide-particularly, a longstanding tension between state administrators in Bogotá and populations in the rest of the country. Throughout the republican history of Colombia, experts in Bogotá, the capital city and administrative center, have attempted to consolidate a national project according to a fragmented vision of the country's territory (Bushnell 1993;Deas 1993;Martínez 2001;Palacios and Safford 2002). These experts have traditionally characterized the Andean, temperate cities, and especially Bogotá, as the spaces of order and civility. Conversely, these same experts have envisioned the vast territories of tropical "warm land" 5 as lawless, mostly vacant, uncivilized, and violent. Margarita Serje (2005) argues that these "cartographies of exclusion" have informed the administration of policies aimed at "taming" these territories through mechanisms like catechization, military pacification, and administrative decentralization.
In every election cycle, these historic fissures surface and manifest in complicated ways. A rich scholarship shows that elections present an opportunity for drug lords and armed groups to violently coerce voters and consolidate their local political power in rural territories and urban peripheries (Romero 2007;Garay Salamanca et al. 2008;López 2010López , 2016. A less studied aspect of Colombian elections, however, is how these territorial histories and social stereotypes also percolate through the concepts and techniques that campaigns deploy to understand voter support, and how political elites ideologically exploit these social stereotypes to justify the permanence of hierarchical, candidate-centered political organizations and the use of coercive methods.
Political experts consider that the opinion/structure vote divide manifests geographically, neatly mapping out to the state's official stratification system. The Constitution of 1991 introduced this stratification system as a progressive measure to determine tax rates, subsidize utilities for poor citizens, and unify the scales used by private utility companies to differentiate their rates. In this system, each municipality classifies its cadastral information according to this six-point scale, with "one" being the poorest and "six" being the wealthiest. Although Colombians broadly use strata talk in daily speech as a shorthand to describe class or income, it is in fact the physical conditions of an area's buildings and the surrounding urban area that determine the stratum, not income.
The strata system provides campaigns and political consultants with a quantifiable language to make sense of voters' behaviors-it delimits measurable demographic populations, data that work well to calculate a candidate's "soft" and "hard" support. These calculations that turn qualities into quantities, what Michel Callon and John Law (2005) term "qualculations, " reproduce and perpetuate deep-rooted social stereotypes and class biases. For example, across the board, political consultants and pollsters advise campaigns to save money by only carrying out public opinion research on and aiming digital marketing tactics toward strata three and higher, whom they consider to be mostly opinion votes. For this portion of the electorate, political consultants and pollsters scope the opinions of different demographic slices, such as by gender, age, occupation, and education level (but not by race because class subsumes race). Conversely, political consultants believe that strata one and two are tethered to specific political brokers and their organizational structure, hence the term "structure vote, " and therefore it is a waste of resources to study their political opinions.
Of course, reality is more complex. I encountered political brokers across strata, and "structure voters" heatedly debating political opinions, and questioning their representatives. However, the opinion/structure vote divide does not tell the story of voters. It tells the story of how political elites easily adapted political marketing methodologies to perpetuate a vertical political layout within a formally participatory institutional framework.
When looking inside the industry's voter management practices, how candidates reach out to voters, and how campaigns impart "discipline, " the neat conceptual division between the opinion/structure vote dissolves. Santos' win over Mockus showed that even with the arrival of new media technologies, old-fashioned pyramidal electoral organization still mattered. My work with Pragmatica reveals that although campaigns generally organize labor and segment campaigning tactics following the social assumptions underlying the opinion/structure vote divide, when it comes to building voters' databases and organizing voter support, campaigns use information technologies to bolster and supervise the work of political brokers in neighborhoods.

Voter management: "Discipline"
After the presidential election of 2010, Pragmatica began working closely with a team of information technology experts who had worked for the Santos campaign under the command of self-proclaimed social media "guru" Ravi Singh. 6 The members of the IT team, mainly self-taught programmers and former hackers, had entered the world of politics when in 2008 they helped organize through nascent social media multitudinous and highly visible marches against FARC in different cities in Colombia and the world. These public demonstrations strengthened then President Uribe's public approval ratings and legitimized his anti-insurgency military campaign.
I met some members of this IT team in person at Pragmatica's headquarters. After working for Santos, they had started their own digital consulting services, a company which I will call Konnect. One of the lead consultants from Pragmatica had arranged a meeting between Konnect and a former client with presidential aspirations interested in cleaning his public reputation, which had been smeared by a recent political scandal. Pragmatica was there to do the public relations legwork, meaning placing stories in major news outlets that would present this politician in a new light. Pragmatica brought in Konnect to spread through social media malignant rumors to discredit the accusations of journalists who had investigated and reported on the politician. The rumors were crafted to suggest that these journalists had ties to the FARC guerrillas, and that they were pawns in their propaganda game (for FARC propaganda, see Fattal 2014). Konnect explained that they would achieve this goal by using a combination of fake social media profiles, a sizable email database, and a multilevel scheme of social media influencers.
After the meeting, I grabbed coffee with Pragmatica's lead consultants, who explained their involvement with Konnect. Usually, Pragmatica buys large email databases and fake social media profiles from these IT experts to spread negative ads and rumors about their opponents. They highlighted that these practices are illegal. Colombian law forbids collecting potential voters' personal information without their explicit consent. Furthermore, electoral regulations prohibit campaigns' use of negative advertisement, and constrain how and when campaigns can use publicity.
Pragmatica usually hires services like the ones provided by Konnect when candidates need to sway the opinion vote-mainly for presidential candidates who already have a political structure across political parties backing them up, but for whom the opinion vote may be decisive. For congressional candidates, and candidates for local elections, Pragmatica advises they build a reliable political structure of voters if they want to stay in office: "An opinion vote candidate may win a seat once, but incumbents have organizations. " Pragmatica reserves other kind of IT services aimed at managing, consolidating, and controlling a political structure. Here, Alejandro-who also worked with the Santos campaign-comes into play.
Alejandro sells to campaigns what he terms "vote surveillance" (vigilancia de votos), a variety of services that range from supporting preelectoral voter registration to measuring the performance of political brokers, called líderes (leaders) in the Colombian campaign lingo. The líderes "sell" voters in their neighborhoods in exchange for personal and collective benefits. A líder will offer to bring a fixed amount of voters to a campaign in exchange for a sum of money, or for a job. But líderes also broker the relationship between candidates and their constituencies and voice the demands of their localities to the representatives they help elect. Yet, even though líderes are the boots on the ground essential to campaigns, candidates, campaign managers, and consultants view them with suspicion. Alejandro has found a niche selling technologies to campaigns aimed at appeasing these suspicions.
The consultants and campaign managers often made disparaging comments about the líderes, and would represent them as con artists in need of strong supervision, of "discipline. " These comments contained a high dosage of classism and racism in line with historic stereotypes that portray the poor as freeloaders, a trait that Colombians call "malicia indígena" ("indigenous malice"). Through Pragmatica, I met an incumbent senator who allowed me to follow his campaign for one day as he toured low-income neighborhoods in Bogotá. The interactions between the campaign managers and the líderes were markedly distant and distrustful. The campaign day started at 7 a.m. and ended at 8 p.m. We went to five neighborhood campaign events organized by local líderes. In each event, regardless of the time of the day, there was food, music, and alcohol. The candidate would come up on stage and deliver the same speech over and over again, step down, shake hands, exchange a few words with attendants, and get back in his car. In appearance, the objective of each event was the speech of the candidate. In reality, the campaign used these events to supervise the líderes.
I rode in the car with two campaign managers. During the time we spent driving, they rarely spoke to me unless I asked specific questions and pushed for an answer. But they didn't speak to each other either, visibly exhausted from the long days on the campaign trail. When we stopped for lunch, one of the managers mentioned to the other manager that "Luz, " the líder from the first neighborhood we visited, was "failing. " What he meant was that she had failed to put together a strong campaign event with many attendants. The managers exchanged looks, a nod, and went back to eating their food in silence. This moment of silence, exasperation, and disapproval completed the picture that this same manager had painted weeks before when I interviewed him. In the interview, the manager emphasized that he possessed an experiential and emotional form of expertise, a form of street smarts, which was very effective in distinguishing whom he could trust and whom he could not. As he put it, "It is a matter of reading the 'vibe' that one gets from people. " He continued to underscore that self-described líderes would sometimes come to the campaign headquarters claiming to have votes, but in reality were only looking for a benefit for themselves without delivering the votes, and that he had learned how to tell real líderes from false ones. But, just in case his vibe-reading failed, the campaign would also deploy other methods to check in on their líderes. The manager continued: líderes have to report to campaigns the names, ID numbers, phone numbers, and addresses of the voters they are bringing in. To ascertain that the names of the voters that the líder reports are correct, the campaign hires a company that pretends to be the Centro Nacional de Consultoría, a reputable public opinion research company, to call those people and verify their voting intention.
The technology that Alejandro sells to campaigns streamlines these alreadyexisting techniques to supervise líderes and to control the voter data that they collect for a campaign, saving campaigns time and money. Alejandro divides his work into three phases: voter registration period, election day, and scrutiny. During the voter registration period (a period of ten months closing two months prior to an election that allows voters to register their ID at a polling post close to their residency address), líderes like "Luz" report to campaigns the information of voters. Alejandro compares this information to the official electoral census released by the Registraduría Nacional del Estado Civil, Colombia's central citizen registry office. It is important to note that the Registraduría releases the electoral census as a tool for individual citizens to check at which polling post they are registered to vote based on their ID number. Alejandro created a bot (an application that performs an automated task) to check quickly whether the voters reported by the líder are actually registered to vote or not. He also hacks the official electoral census file released in DVD format by the Registraduría to access the information linked to each ID number. This information includes the state (Departamento), city, neighborhood, polling post number, and the exact polling post table where each ID number is registered to vote.
Based on this information, campaigns know how many voters to expect from each líder, they make sure that voters are registered to vote near their homes (Colombian elections take place on Sundays), and they organize the logistics for election day, requiring líderes to coordinate the operation in their own neighborhoods. Based on this information, campaigns also choose testigos electorales, loosely translated as electoral witnesses. Although originally envisioned as a tool to widen participation and prevent electoral fraud, campaigns hire testigos electorales to oversee voters. An electoral reform in 1979 allowed political parties to send representatives to oversee elections inside the polling posts, and to make sure that table judges count and record votes for a party correctly. The electoral witnesses can ask for vote recount on site, and they report in writing anomalies to campaign organizations, which, in turn, can file formal complaints the next day. Specific campaigns hire electoral witnesses to guard their votes. Campaigns deploy electoral witnesses in polling posts across the six strata, but mostly emphasizing neighborhoods where they expect a high influx of "structure votes. " Alejandro trains the electoral witnesses to recognize electoral fraud, and to properly record the information contained in the E-14 form, an official, publicly available form where polling table judges record voting results. The electoral witnesses oversee two to three tables, and then take the information to the polling post manager (hired by the campaign), who uploads the information to a centralized system managed by Alejandro using software that the latter developed. However, Alejandro complains that the electoral witnesses often do not record the information properly, and that he often does not have complete data to supervise the votes of the campaign during the early stages of scrutiny. To oversee the votes of the campaign, he accesses the plain text file released by the Registraduría by midnight of election day with all the results (election reports are remarkably efficient in Colombia). This file gathers the number of votes per candidate, per polling post, and per voting table in each city and in each municipality. Although this file does not contain the ID numbers of voters, using the voter information initially reported by the líderes, and comparing them to the results for the candidate by voting table, Alejandro knows which líder delivered. Based on this information, campaigns may reward a líder, or stop working with that person in the next election cycle.
A campaign's indictment of a lider's effectiveness may end that person's political career. Líderes who work for different political organizations and for different parties usually know each other. I visited a recently elected house representative in his office in Congress. Outside his door stood a line of líderes coming to voice the concerns of their neighborhoods and also to receive their pay. I chitchatted with some of them while they waited. To my surprise, many of them knew the campaign manager for the senator, even though the representative and the senator belonged to different parties, and they also knew what type of job this manager had landed in exchange for his services. Through another líder, Mrs. H., who belonged to the same organization as these other líderes, I learned that líderes often cross party lines. Mrs. H., who works in a middle-class neighborhood in Bogotá, switches liberal-leaning parties-because of her family's traditional ties to the Liberal Partydepending on who will give a job to her son. Gossip of ineffective líderes spreads quickly in these networks, crossing party lines.
For Alejandro, however, gossip is not enough. He aspires to one day develop a qualification system for líderes based on their voter turnout records that campaigns can use to decide from the outset whether to work with a specific líder. He would, however, need líderes to keep good records of their work. Alejandro hopes that campaigns someday will collect enough voter data to effectively supervise líderes without depending on the information they provide.
Alejandro's work exemplifies the Colombian version of what in the United States has been termed "microtargeting, " a data-intensive voter management practice. In the United States, parties collect detailed information about voters, including voter rolls and consumer reports, to microtarget the message of campaigns to potential voters using techniques like direct mail, or recruiting volunteers to knock on doors. The objective of microtargeting is to organize get-out-the-vote efforts. Scholars have observed that microtargeting often uses forms of social pressure to persuade individual voters (Issenberg 2012), it fragments political debate by adapting a campaign's message to resonate with the life experience of voters instead of confronting their biases (Abrajano 2010), and it overly scripts citizens' participation in campaigns (Nielsen 2012).
The Colombian case shows that data-intensive campaigning may also contribute to perpetuate vertical political organizations positioned to trade voters' information. More subtly, Colombian political consulting illustrates that voter management technologies designed in accordance with stereotypical notions of voter behavior, and created to bolster on-the-ground political networks, produce "information that self-describes itself " (Corsín Jiménez 2011: 184-85), meaning unverifiable information that precludes external participation, and that relies on its own terms of verification. One has to trust that the opinion/structure vote divide accurately describes voters, and that every level of these pyramidal electoral organizations reports the information correctly in order to believe that the voter management techniques that professional political consultants sell to campaigns will determine a candidate's win. Like Alejandro, who blames the líderes for working with faulty information, Pragmatica's consultants often explain candidates' defeat by blaming external conditions of the campaign, such as not raising enough money to support the work of a political organization. These self-contained or self-actualizing voter information systems contribute to reproduce Colombia's traditionally atomized political sphere, and the power of strongmen. Jennifer Schirmer (2008: 220) insightfully observes that Colombia "has strong and vibrant intrasectoral debate, yet intersectoral silence, " referring to how groups with different political ideologies and from different social backgrounds rarely find opportunities to debate. I suspect that the concentration of information and access to voters in pyramidal political organizations may also contribute to accentuate intersectoral silence, and, conversely, candidates' amplified voice.
Alberto Corsín Jiménez (2011: 187) further claims that "information is a blaming-resource used by society to reallocate its funds of responsibility, " turning corporate information systems into "systems of moral accountability" that dislocate trust and risk from social relationships. When information systems strip the trust out of political relations, and centralize this trust in systems that service a candidate's political ambitions and public persona (Lempert and Silverstein 2012), political practice and political truth become inextricably linked to a candidate's image. To conclude, let's turn to Álvaro Uribe Vélez, the ultimate cyborg political machine.

Conclusion: Álvaro Uribe Vélez
It is not a coincidence that Alejandro had worked in the public demonstrations against FARC that bolstered Uribe's popularity during his presidency. Uribe's government effectively used media technologies, marketing, and traditional political networks to amplify Uribe's message, and to build a robust political base that legitimized his militaristic government.
Uribe cast over Colombia what has been described as an "authoritarian spell" (Hernández Paredes 2003), in which people willfully embraced restrictions to their own rights to preserve security and order. Uribe's government slogan, "Seguridad Democrática, " exploited people's fears about FARC and a failing economy. Uribe redefined the meaning of security to mean both personal safety and economic prosperity. Breaking away from previous governments' understanding of security, in Uribe's formulation security would not be the product of a peace negotiation, but it would only come after the military defeat of insurgent groups. Throughout his presidency, he justified authoritarian modes of government as necessary to preserve democracy, and he used measurable public opinion support to legitimate the use of coercive tactics (from military action to public defamation techniques) against those who openly opposed him (Roldán 2010).
For the sake of democracy and maintaining order, every Saturday, Uribe would travel with his cabinet, the police, and sometimes the military to the provinces to host televised town hall meetings, known as consejos comunitarios (community councils), where he would listen to citizens' concerns for many hours. In June 2007, I went to one of President Uribe's community councils in the Caribbean city of Santa Marta-only after my uncle, a local politician, pulled some strings to get me on "the system, " meaning the event's list. The attendees were mainly local bureaucrats, local journalists, and public officials from Bogotá, "people who live off politics, " as my uncle explained.
A frenzied audience greeted the president. Once people settled down, Uribe and his cabinet sat on the main stage to discuss different issues of concern for the region. Acting as the moderator of the meeting, and deciding who could speak, Uribe asked the ministers, one by one, to report on the status of infrastructure projects in the region. After the reports, Uribe encouraged the audience to voice concerns and ask questions. People stood up and grabbed the microphone to ask the president for favors, such as building a gym in a public school or paving a street in their municipality. Uribe would urge the appropriate public official to take note.
Shortly after listening to attendees' grievances, the meeting reached its climax when the conversation shifted to microloans. The people sitting in front of me stood up. They all wore the same t-shirt bearing an advertisement for the government's microloan program managed by the Banco Agrario (Agrarian Bank). The president introduced them as the beneficiaries of microloans who would receive their checks during the community council. Before giving away the loans, Uribe asked for a priest. One came to the center stage and blessed the checks.
One could dismiss the community councils as pure populist spectacle. However, they were incredibly effective in consolidating Uribe's coalitions in the regions. The televised community councils showed politics in the making-how Uribe used his position to funnel state resources to those politically aligned with him. The fact that to attend the councils and to register in the "system" I had to deploy personal connections with political insiders is quite telling of these bodies' coalition-building function.
The novelty of the councils, however, is that these coalitions also came together to further propaganda. The councils drilled in people's minds the message that "Seguridad Democrática, " Uribe's militaristic machine, which routinely violated human rights, also brought prosperity to the regions. To this day, Uribe's government message continues to resonate with many Colombians who fervently support him, and Uribe continues to capitalize upon their support.
In 2016, President Santos promoted a referendum to approve the Peace Accord resulting from four years of negotiations between his government and the FARC guerrillas to end the longest guerrilla war in the Western hemisphere. In an ironic twist, this time round, liberals from the cities, including Mockus (Santos' opponent in 2010), strongly backed Santos. However, former ally, and now leader of the opposition, Senator Álvaro Uribe played a central role in galvanizing conservative groups to reject the Peace Accord, and won.
Promoters of "No" effectively convinced a conservative majority of voters that the Peace Accord would turn Colombia into a socialist, atheist country. The promoters of "No" used social media platforms, existing local political networks, evangelical and Catholic pulpits, and old-fashioned street advertisement to promote their message. Juan Carlos Vélez, the campaign manager of "No, " revealed in a controversial interview that the campaign spun its message to speak to different voter segments and to anger them.
"No" misrepresented the gender emphasis of the Peace Accord to instigate homophobic sentiments among religious voters. The campaign claimed that the government would impose what they called a non-Christian "gender ideology" in schools. "No" also distorted the victim reparations portion of the Peace Accord to scare business owners that the government would seize their property, and to convince senior citizens that the government would slash their retirement savings to fund the implementation of the agreement.
Much of "No's" success selling these distortions depended on Uribe's experience with many of these communities, and his ability to speak to them effectively-conservatives believed him because they had seen the president on TV building infrastructure, being a good Christian, and handing down the benefits of a militaristic state. Uribe had credibly embodied conservative values and aspirations, and the televised community councils served him to consolidate long-lasting regional alliances with conservatives by bringing the symbolic magic of TV to seal these pacts.
Although the government decided to move forward with the Peace Accord despite the fact that "No" won in the polls, now the fate of the agreement is tied to the elections of 2018. Uribe tested in the referendum an antigovernment, antipeace, religious electoral platform that proved to be effective. Heading toward the congressional and presidential elections of 2018, Uribe and his party are seizing every opportunity, like connecting the Venezuelan crisis to the Peace Accord, to campaign against the Colombian government.
If by the late twentieth century we all were "chimeras, theorized and fabricated hybrids of machine and organism" (Haraway 1990: 150), by the twenty-first century we have fabricated cyborg leviathans, hybrids of image and sociality, that have become our politics. Uribe's success depended on bringing local politics and political marketing in close alignment-on running an effective cyborg political machine. I suspect that Uribe is part of a trend, and that similar types of machines may be fueling the rise of populism in Northwestern democracies.
This essay has argued that political consultants use information technologies to boost local political operations. I have shown that Colombian political elites introduced political marketing to modernize campaigning and take power away from political brokers. Yet political marketing's vertical management practices easily translated to, and served political elites to perpetuate, Colombia's traditional authoritarian political formats. Arguing against seeing authoritarianism and antidemocratic impulses as native attributes of democracies in the developing world, I instead have shown that Colombian political elites found in political marketing a practical handbook to preserve authoritarian local powers in the wake of democratic reform. The case of Colombia suggests that information-centered, professionally managed campaigns rely on the work of established political networks that in return also consolidate their power through these operations. The rise of extremism, authoritarianism, and hyperbole in mainstream politics may be the residual product of attaining the cyborg political machine.
interviews, and my informants, who generously shared their work. Lastly, my gratitude to Dominic Boyer and James Faubion, who believed in this project when it first took shape at Rice University, and to MIT Anthropology for its continuous support. My appreciation to MIT's Open Access Fund for supporting the publication of this article.