In surveillance capitalism (Zuboff, 2015), surveillance is often monetised to the detriment of the masses in order to benefit the few and exacerbate existing inequalities. Digital forms of surveillance are part of our everyday life and we cannot opt out of them easily as surveillance has become ingrained in smart cities, automated mobility and e-government services, in work places and consumption routines, in payment systems and entertainment. Since the world of massive information infrastructure underpins all social processes, it is impossible to escape the surveillance gaze. The economic rationale of ‘digitising in order to increase efficiency’ is being progressively enforced on a legal basis, for instance when attempts to avoid facial recognition technology in public space by covering one’s face is legally labelled ‘disorderly behaviour’ (Malik, 2019). This represents a fundamental turn from ‘innocent until proven guilty’ as avoidance of surveillance itself becomes a transgressive act.

The use of surveillance technologies to reverse the surveillance gaze has gained certain traction in theorising ‘the gaze from below’. Authors have made significant contributions to the other side of surveillance and the potential thereof by attempting to answer a number of quintessential questions: How can individual autonomy and dignity, fairness and due process, community cooperation, empowerment and social equality benefit from surveillance? What are the ways in which those who are subject to surveillance manage, negotiate and resist the spread of surveillance? The ‘new surveillance’ (as defined by Garry Marx, 2002) is forming complex networks of power relations and resistances (Green, 1999). Objects of surveillance have become empowered agents, not the passive and powerless parties they used to be (Ball and Webster, 2003). They are growing into active entities that reflect on surveillance practices and try to escape and avoid them or draw public attention to illegitimate, discriminatory or otherwise unfair practices. Resistance and counter-measures were labelled and conceived very differently in the literature, for example, as ‘counter-surveillance’ (Monahan, 2006), ‘surveillance and empowerment’ (Monahan et al., 2010), ‘inverse surveillance’, ‘sous-veillance’ (Mann et al., 2002) and the ‘hijacking’ of surveillance (Koskela, 2006, 2009). Mann et al. (2002) deem ‘sous-veillance’ to occur when objects of surveillance use oppressive tools against the oppressors themselves. According to Monahan (2006), ‘counter-surveillance’ as a form of tactical interference often results in surveillance technologies working against themselves, in order to redress institutional power asymmetries.

The assessments of the power of the ‘gaze from below’ have been mixed: Andrejevic (2007) wrote about ‘lateral surveillance’ in his seminal book iSpy and claims that the use of surveillance technologies replicates the underlying logic of the ‘all-seeing’ or ‘sensor’ society and produces a flood of images that the information technology (IT) industry encourages. The trend of peer-monitoring networks has amplified and replicated the power of government or corporate surveillance in a climate of perceived risk and savvy scepticism (Andrejevic, 2005). On the other hand, lateral surveillance can have an empowering potential. Koskela (2006), for example, claims that interactive webcams and citycams, whose detailed real-time streaming video is used to a promote space, have empowering roles.

The development of information and communications technology (ICT) coupled with the rise of big data and algorithms brought the promise of unprecedented citizen empowerment through both active engagement as well as passive oversight of the governance of the state. The new technologies were supposed to return power to those to whom it belongs in a democratic society – the citizens (Gigler and Bailur, 2014). However, reality has failed to live up to that promise as it is becoming increasingly clear that the mere application of new technologically supported tools does not stimulate citizen engagement by itself, at least not in all parts of the society (at the same time). The digital divide between the underprivileged members of society – whether individuals, groups or entire countries – that do not have access to computers and consequently the Internet, and the wealthier, privileged members that fully enjoy the opportunities ICT offers has been widely discussed (Compaine, 2001; Norris, 2001; van Dijk, 2006; Warschauer, 2004). The emergence of big data and artificial intelligence (AI) has created a further divide among the technological ‘haves’ as only a fraction of those with access to computers also have access to databases, processing power and storage, on the one hand, and the knowledge and expertise to put them to use, on the other. This has created a new class described as ‘big data rich’ (as distinct from ‘big data poor’) (Boyd and Crawford, 2012), which has monopolised these powerful tools for decision-making and prediction to the detriment of the majority of the population. The statement of Google executive and Egyptian activist Wael Ghonim in the aftermath of the Egyptian revolution that, ‘If you want to liberate a society, just give them the Internet’, no longer holds true, as regular citizens cannot comprehend the sheer amount of data it now provides (Rao, 2011).

As the trend towards the paradigm of ‘open data’ gains traction, it becomes necessary to ask how much emancipatory potential this paradigm actually carries. Since the public sector generates a large amount of data, an important prerequisite to bridging the existing data divide is the availability of open data, that is (governmental) data that anybody can access, use and share. Numerous stakeholders on both national (data.gov, data.gov.uk, govdata.de, dati.gov.it, podatki.gov.si) and international levels (EU Open Data Portal, UNdata, World Bank Open Data) have in recent years made their data sets freely available online with the aim of ensuring public transparency and accountability, stimulating democracy and citizen participation and improving public sector efficiency.

It is not enough that the data are only available; they also need to be meaningfully organised and presented in order to encourage citizen participation. In this process of extracting meaning for various social ends, infomediaries (i.e. third parties that analyse and visualise open data made available by the government) are becoming more important than individuals who usually have neither the knowledge nor the capabilities to transform the ‘raw’ data into understandable and socially meaningful actionable information that produces knowledge. Mayer-Schönberger and Zappia (2011) illustratively ask whether we ‘will see the rise of a new caste of intermediaries that hold the key to making sense of the seas of data now accessible’. Infomediaries are gaining power over the data and the way in which they visualise, map and present them to citizens may shape the way they perceive them. The question then is how much power ‘the new caste’ absorbs and whether the new ‘citizen data scientist tools’ (CDSTs) can democratise surveillance capitalism.

In ‘Four Critiques of Open Data Initiatives’, Kitchin (2013) concludes that ‘we lack detailed case studies of open data projects in action, the assemblages surrounding and shaping them, and the messy, contingent and relational ways in which they unfold’. This chapter provides an informative analysis of specific CDSTs in a specific post-socialist country. The open data initiatives that have triggered the rise of the citizen data scientist (with the exception of Zlovenija, which uses Facebook users’ posts) have much to offer, but we ask who possesses the data, what ends their analysis and visualisations serve and who is empowered by them.

In this chapter we present several CDSTs used in social practice in Slovenia and evaluate them according to common criteria: the type of tool (e.g. non-governmental organisation (NGO) based, state based, intended for the implementation of an international initiative); who or what are the targets and end goals of the tool; datafication intensity (the use and reuse of data, what the data sources of the given tool are); the type of resistance (the force of the tool, ranging from passive avoidance to active subversion); and the empowerment dimension (who is empowered). The concluding section provides an analysis from a more nuanced perspective of the empowerment of CDSTs. Methodologically, the chapter is based on a review of many more CDSTs than presented herein and that are up and running in Slovenia and beyond, but we decided to narrow our focus to locally informed CDSTs in order to reveal context-specific uses of data and their potential to empower users. The findings from the review were double-checked and substantiated in semi-structured interviews conducted with the anonymous author of Zlovenija and with Matej Kovačič, one of the designers of Supervizor (subsequently Erar), the head of analytics and information security at the Commission for the Prevention of Corruption of the Republic of Slovenia (CPC) at the time the tool was set up.

Case studies of CDSTs

Erar: monitoring public spending

In 2011 the CPC started a project named (in translation) Transparency, of which Erar,¹ an online application for monitoring the expenses of public bodies, is an important part. The rationale behind the tool is that the transparency of financial flows between the public and private sectors increases the level of responsibility of public office holders for effective and efficient use of public funds, facilitates debate on adopted and planned investments and projects, decreases the risk of illicit management and abuse of office and above all limits systemic corruption, unfair competition and clientage in public procurement procedures (Kovačič, 2016). Erar’s predecessor, Supervizor (Supervisor), was awarded a ‘United Nations Public Service Award’ in 2013.

The online application enables simple browsing through the financial transactions of public sector bodies, government spending and a graphical presentation thereof. Users can view all monetary transfers from a selected budget user, for example an agency or public primary school, or even all money transfers from a budget user to a selected company. Data can also be presented for a specified period of time. The application also shows data about public procurements and business entities in Slovenia along with their management and ownership structure and information from their annual reports. An important part of the application is a subsection that presents a list of publicly owned companies and information about them (see Figures 8.1 and 8.2).

About data sets

Several authorities provide nine types of ‘open-data’ data sets, on which the application is based:

Some types of data are not incorporated in the application, such as transactions where the purpose is the payment of a union membership or transactions made by intelligence agencies.

Use cases

Erar can be used for several investigative purposes and – despite being the tool of a government agency – it has contributed to the substantial public and political outrage since its inception. At the municipal level, the use of the tool has enabled the discovery of several misappropriations made by accountants in municipalities. At the state level, other agencies such as tax authorities have admitted that the tool has enabled them to perform investigations more effectively. It has also helped trigger government change.

In 2013 the CPC issued a report with the claim that a ‘strain of corruption risk’ existed within the sphere of the then prime minister Janez Janša. By using the Supervizor tool and merging data, the Commission identified the prime minister’s significant wealth of unexplainable origin, which led three coalition partners to leave the government. A majority of the members of the parliament then passed a vote of no confidence and the government was subsequently dismissed on 27 February 2013. The collapse of the government was not a consequence of a criminal conviction, or sole use of the Supervizor tool. Rather, it was a result of the new idea taken up by the CPC leadership, with Goran Klemenčič as president of the CPC, that public data should be more extensively correlated in order to prevent corruption among the highest public officials. Moreover, the tool triggered direct changes at the ministry level. When in 2015 Supervizor published data on budgetary payments to individuals that were paid in addition to their salaries, it revealed that relatively large amounts were paid to the elite of the higher education system. In 11 years, more than 1 billion euros of budgetary funds were spent on such additional payments to the educational elite, representing more than one-tenth of Slovenia’s annual budget. The then minister of education, science and sport, Stanka Setnikar Cankar, was forced to resign as it was revealed that as the former dean of the Faculty of Administration of the University of Ljubljana, she earned on average EUR 4,000 per month in addition to her salary, totalling EUR 636,000 in 11 years. However, the legal battle between her and the CPC due to irregularities concerning the publication of the data resulting in a violation of her privacy rights is not yet finished.

The Erar tool enables the detection of direct violations of restrictions on business activities due to a conflict of interest (Article 35 of the Integrity and Prevention of Corruption Act 2010). Such a conflict exists when a public sector body or organization that is obliged to conduct a public procurement procedure in accordance with the regulations or that carries out a procedure for granting a concession or other form of public–private partnership (such as ordering goods, services or construction work), enters into a public–private partnership or grants a special or exclusive right to an entity due to specific personal ties. The prohibition enters into effect when a public official or a family member thereof participates as a manager, member of management or legal representative in the company at issue, or has more than a 5 per cent level of participation in the founder’s rights, management or capital. Contracts awarded in breach of Article 35 are void. Public bodies have to report potential restrictions of the business activities of their office holders by means of an online form within one month of assuming office and no later than eight days after any change thereto.

When it entered into operation, Erar enabled the swift detection of violations of restrictions of business activities as links were established between the database of restrictions of activities and the database of payments from public bodies (in order to detect prohibited business activities) and the Slovenian Business Registry (in order to detect which officials should have reported business restrictions but failed to do so). Kovačič (2016) compiled a list of suspected violations of restrictions on business activities in 2012 and showed that violations took place in 68 cases (429 contracts) and the total value of illicit business amounted to EUR 1,436,208.28. The CPC repeated the analysis and no violations were found a year later (Kovačič, 2016).

Another use of Erar is to compare the public procurement success of companies in periods with different governments: Is there a link between individual governments and the disbursement of funds to particular companies? Or, to put it differently, are companies significantly more successful in public procurement procedures during a particular government, and vice versa? Kovačič (2016) showed that there are significant differences not only in the success of particular companies, but also with regard to particularly vulnerable sectors. ICT companies providing equipment to public authorities multiply their profits under one government while their business comes to a halt altogether when the government changes (see Figures 8.3 and 8.4). Such high inflexibility of the market was established especially regarding ICT services, pharmaceutical products and construction work. Some companies are highly dependent on financial transfers from direct budget users as they receive a great amount of their income only from budget users, which constitutes a noticeable risk of corruption (Kovačič, 2016).

Who Influences? (Kdo vpliva?): monitoring lobbying

The platform Who Influences?² was devised to provide insight into lobbying processes in Slovenia through visualisations of two different kinds of networks: (1) the network of contacts between private enterprises and their lobbyists, on one hand, and politicians, political parties and governmental institutions, on the other; and (2) the network of transactions between companies in which lobbyists are legal representatives. The online visualisation shows a network of interconnected dots of various sizes, in which every dot represents either a lobbyist, an organisation or a company, a lobbied institution (e.g. the Government of the Republic of Slovenia, a ministry, the National Assembly or a local community) or an individual decision-maker. The size of the dot increases with more lobbying contacts – the more active a participant is, the bigger the dot becomes. Clicking on a dot provides the user with further information about the time, location and purpose of the lobbying activity and organisations that have established connections with a particular decision-maker (see Figure 8.5).

The platform Who Influences? is based on publicly available data on lobbyists and lobbying activity that have to be reported to the CPC on the basis of the Integrity and Prevention of Corruption Act. Consequently, its quality and accuracy depend entirely on self-reported data, which in the case of some ministries and municipalities is lacking or even completely non-existent. It is possible that some areas of government are not interesting for lobbyists at all; however, it is far more likely that some partakers fail to comply with the requirements of the law and do not report lobbying activities to the Commission.

Parlameter: monitoring parliamentary activities

Parlameter is a parliamentary informatics tool that ‘facilitates monitoring of the National Assembly’s working process through an analysis of voting and transcripts’.³ Similar digital visualisation tools have been developed to provide parliamentary oversight in numerous countries worldwide as well as on the international level (e.g. Vote Watch Europe monitors the activity of the European Parliament). Parlameter therefore follows the growing trend in parliamentary monitoring known as ‘parliamentary informatics’, that is, the application of ICT to monitor parliamentary work and legislative procedure. A recent study showed that among 191 parliamentary monitoring organisations (PMOs) monitoring more than 80 parliamentary bodies, including national parliaments, sub-national parliaments and international legislative bodies such as the European Parliament and the General Assembly of the United Nations, approximately 40 per cent use parliamentary informatics and in most cases ICT technology is employed to provide a user-friendly visualisation tool (Mandelbaum, 2011).

Parlameter provides a user-friendly instrument for exploring the legislative procedure by searching for information about: (1) individual members of the National Assembly (hereafter MPs (Members of Parliament)); (2) the political parties represented in the National Assembly; (3) plenary and extraordinary sessions of the National Assembly; and (4) legislative proposals. First, Parlameter enables monitoring of individual MPs, a feature it shares with a majority of other PMOs, since it promotes accountability and more effective public scrutiny (Granickas, 2013; Mandelbaum, 2011). Apart from MPs’ demographic data (age, education, previous employment, number of terms of office, party affiliation) and links to their social media sites, Parlameter also provides indicators of MPs’ (in)activity in the legislative procedure, for example MPs’ attendance records, committee memberships and the number of parliamentary questions posed and speeches given per parliamentary session. Users may also view how each MP has voted on a particular legislative proposal, while comparison tools enable users to easily identify MPs with similar or different voting preferences. Additionally, Parlameter analyses parliamentary discourse, which significantly distinguishes it from other PMO tools. It does not just facilitate access to MPs’ speeches during both plenary and committee sessions, but it identifies the words most commonly used by a particular MP and positions him or her according to the diversity of his vocabulary used in parliamentary discourse. The tool also conducts a stylistic analysis of the language used in parliamentary discourse by measuring three features: the use of euphemisms, colloquialisms and stylistically marked words.

Second, Parlameter provides a very similar set of information on the political parties represented in the National Assembly: general information about the party (the leader of the parliamentary group and his or her deputy, the list of elected MPs); the voting records of the parliamentary group; the number of parliamentary questions posed and proposed amendments; and an analysis of parliamentary discourse, including transcripts of all speeches, the most commonly used words, the diversity of language and a stylistic analysis.

Third, the subsection on the legislative process provides information about proposals that went through the legislative procedure, which is accompanied by a short summary, voting on amendments and the legislative proposal as a whole and the current status of the proposal (adopted, rejected or still in the legislative procedure).

Additionally, Parlameter provides a variety of tools: ‘Parliamentary Compass’, which positions all MPs in a two-dimensional virtual space based on their compatibility of views (MPs with similar views are placed closer together); ‘Groups of Words’ shows how often an MP or a party mentions a particular word or group of words during the sessions of the National Assembly; and ‘Parlameter Notifications’ sends a notification to users via email when a selected word is used in the National Assembly. Parlameter’s newest feature is ‘Commentality’, a tool designed to measure public opinion by sending feedback, similar to liking on social media sites, regarding statements referring to legislative proposals (see Figure 8.6). Its aim is to keep the audience engaged without enabling users to comment on a particular proposal (Commentality, 2019).

To sum up, Parlameter is primarily a convenient tool for researchers and journalists, but at the same time its goal is more far-reaching: it pursues the aim of making the legislative procedure of the National Assembly more transparent, accessible and therefore understandable to the general public (Gligorović, 2016). When Parlameter was introduced in December 2016, the president of the National Assembly, Milan Brglez, noted that ‘more intensive and user-friendly familiarisation with the work of the National Assembly is a necessary prerequisite for a better understanding of its work … Such tools enable inclusive democratic oversight and present the potential for more diverse and intense political participation’ (cited in Gligorović, 2016).

Although information on the working process of the Slovenian National Assembly is publicly available on its official website, it is published in a form that is difficult for a regular user to effectively use, which severely complicates or even undermines parliamentary monitoring by civil society. Parlameter, on the other hand, provides information in an accessible, understandable and searchable format. It is the new infomediary for extracting meaning from data for the wider public.

Legislative Activity Violation Counter

The Resolution on Legislative Regulation (2009) establishes principles regarding the legislative procedure, guidelines for the assessment of the effects of new laws and guidelines for consultations with relevant stakeholders in the course of preparing legislative proposals. The Resolution targets members of the National Assembly and seeks to increase the overall quality of legislation. Inter alia, the Resolution also determines minimum standards for consultations with the public by establishing a minimum threshold for consultations ranging from 30 to 60 days.

In order to test the effectiveness of the Resolution, the national NGO umbrella network Centre for Information Service, Co-Operation and Development of NGOs (CNVOS) designed the Legislative Activity Violation Counter (CNVOS, n. d.) to measure how often the Resolution’s public consultation provisions are violated. The Counter (see Figure 8.7) covers all the regulations for which the Resolution determines a minimum period of public consultation and all other acts for which such consultation is envisaged by the Rules of Procedure of the Government of the Republic of Slovenia.

The Counter’s website collects data on the total number of violations and presents these by individual ministry, with an indication of the names of the regulations for which the prescribed minimum period for public consultation was not observed and the number of deviations. The website also shows a list of proposed regulations with an unspecified deadline for the submission of comments and those proposed regulations without any invitation to submit comments where a draft regulation was simply published on the website of a ministry.

Žvižgač.si: the Slovenian whistle-blower portal

‘Žvižgač’⁴ is an online portal that enables users to submit public interest information anonymously and safely, without fear of being exposed to their family, friends, employers and the media. Since in most countries (including Slovenia) whistle-blower protection is still not regulated legislatively, individuals that decide to speak up and expose corruption, fraud or other cases of misconduct risk numerous negative consequences, including public humiliation fear of retaliation, job loss or even criminal prosecution. Protection of their identity is at this point the only way to avoid undesirable negative impacts due to whistle-blowing. With this aim in mind, various open-source software applications have been developed since around 2018, for example, GlobaLeaks, AdLeaks and SecureDrop. Žvižgač is based on the latter, which was created by Aaron Schwarz and has now been adopted by established international media (e.g. Forbes, the Guardian, Financial Times, New York Times).⁵ Before being disclosed to the wider public, all information and tip-offs transmitted to the portal are independently verified and assessed as to whether the publication is in the public interest (Žvižgač, 2019). Potential whistle-blowers may also find useful tips on how to use Tor software or how to legally safeguard their position on the Žvižgač website, for which it provides a comprehensive tool for informing and combatting corruption or similar illegal or undesirable practices.

iSee: paths of least surveillance

iSee is a web-based and wireless application based on a database updated in real time that allows any user to programme urban routes with the smallest exposure to surveillance cameras (IAA, n.d.). Developed by the Institute for Applied Autonomy (IAA), iSee charts the locations of closed-circuit television (CCTV) surveillance cameras in Ljubljana, Slovenia (as well as in New York and Amsterdam, the Netherlands) in order for users to find routes devoid of public space surveillance or with the least surveillance. The programme uses starting and destination points and calculates the shortest distance without (or with the fewest) security cameras (see Figure 8.8).

The idea behind iSee, as explained by its designers, is as follows:

While the aim of the application is not to directly interfere with the surveillance apparatus of cities, the project was launched in the aftermath of the September 11 attacks and was extremely pertinent as to raising awareness of the increasing overreaction of security apparatuses around the globe.

Mapping the public space without surveillance was a realistic and practical alternative in the early 2000s in Ljubljana, but with the proliferation of the monitoring of public space in the 2010s, when the number of cameras skyrocketed,⁶ it has become impossible to avoid the prying eyes of CCTV cameras and users can now only find routes with the shortest distance and the fewest security cameras.

The iSee project triggers pertinent question related to ‘renegotiating the limits of private and public domains’ (Nakashima et al., 2010, p. 295). The proliferation of surveillance of public space with CCTV cameras (increasingly augmented with gunshot detection systems, or face, gate and other similar biometric recognition systems) triggers a dilemma associated with the traditional view that a person cannot reasonably expect privacy in public space. iSee reflects discomfort with blurring the lines of anonymity in public space in a similar vein as mosaic theory. The latter builds on the insight that when combined with other data, information about a person becomes much more telling and suggests that an expectation of privacy exists (or at least should exist) even in public space, due to the way people live their lives in the digital age (Završnik and Križnar, 2018). While not radical in its approach – compared to other projects that aim at destroying cameras⁷ – the project nevertheless renders visible the larger pattern of surveillance proliferation and calls into question its purpose, agenda and effects (Monahan, 2006).

Redaar: traffic-ticket monitoring

The traffic-ticket monitoring mobile app Redaar, freely available at GooglePlay, maps past infractions and helps drivers avoid parking in locations where and at times when they are more likely to be ticketed. Slightly fewer than a million traffic violations occurred between the beginning of 2012 and September 2014. Given that there are 1,300,000 registered vehicles and 1,400,000 active driving licences in Slovenia, this is a relatively high number (Plahuta, 2015). The great majority of the traffic violations were parking and toll tickets, which shows how they have become a major source of revenue for municipalities.

The idea of the app is to simultaneously monitor minor offence authority activities and empower users to escape enforcement for parking violations. Drivers can set the app to show the following issued traffic citations: parking, speeding, driving while using a mobile phone, ignoring safety belt laws, unpaid tolls, driving under the influence (DUI) and traffic accidents (Virostatiq, 2015). The app, created by data scientist Marko Plahuta, is based on the analysis and visualisation of two years of traffic violations data. Data sources were obtained from: the traffic warden service in Ljubljana; the national police; the Motorway Company of the Republic of Slovenia (DARS d.d.), which manages state roads; the public undertaking ‘Ljubljanska parkirišča’, which manages car parks in the capital; and the local traffic warden services of several Slovenian towns. Given the temporal data for each issued ticket, the app shows on which streets drivers are more likely to be ticketed in the morning, at midday or in the evening (see Figure 8.9).

Similar applications for mapping traffic violations are focused on crowdsourced speed-camera alerts (e.g. in the Slovenian context, ‘BrezKazni.si’ and ‘Radarji in Hitrostne kamere’, available at GooglePlay). For instance, the speed-camera mobile app Traffic Light Cameras (designed by bigDream) maps fixed and mobile speed cameras and includes extensive country-specific information on fines for speeding.

‘Zlovenija’: policing ‘from below’ or vigilantism?

Zlovenija⁸ was a response of civil society to the hatred directed at migrants in public posts on Facebook that emerged at the peak of the migrant crisis in 2015. The idea was simple: the most brutal, extreme, outrageous, vile comments posted on Facebook were exposed on a Tumblr page and supplemented with the name and enlarged profile picture of the person posting them. The authors of Facebook comments could request the removal of the post if they deleted the original comment and apologised. The apologies subsequently replaced reposted Facebook posts on Zlovenija. Tumblr is a visual microblogging website that is easy to use and crowdsource new content, thus it provided a perfect platform for displaying the faces of hatred to the widest audience possible (Oblak Črnič, 2017). This initiative, which started as a spontaneous response to the hatred infecting the country, became a powerful tool in the hands of civil society in a matter of just two weeks (Interview with Zlovenija, 2018). What was at first just an overlooked Tumblr page became a pillory with the mission of initiating discussion on the issue of hate speech, reminding people that the Internet is a public space and that words have meaning and consequences, holding up a mirror to society and condemning all intolerance and violence (Zlovenija, 2015). Zlovenija triggered mixed responses. On the one hand, a part of the general public agreed with and even admired the objectives Zlovenija was trying to achieve. On the other hand, it was condemned for being too extreme, no better than what it was trying to expose, and comments often raised questions as to both its legality and legitimacy.

The T-shirt in Figure 8.10 reads (in translation): ‘Run. Do good’, followed by a comment stating: ‘On to trains. That is, freight trains and directly to Dachau’ (referring to the concentration camp Dachau) (Zlovenija, 2015).

Zlovenija differentiates from the other visualisation tools mentioned since it was created by civil society not to monitor those in power, but to expose immoral (if not illegal) acts of members of society. Therefore, it cannot be understood as a form of either sous-veillance or counter-surveillance, but rather as an act of digital vigilantism, a practice described by Trottier (2017, p. 56) as ‘a process where citizens are collectively offended by other citizen activity, and respond through coordinated retaliation on digital media’. However, the phenomenon of Zlovenija shares numerous characteristics with sous-veillance techniques as they (1) use the technology to hold up a mirror to society; (2) put an empowerment tool in the hands of civil society enabling active participation; and (3) endanger the right to privacy of the exposed individuals. All of the above-described sous-veillance tools are based on publicly available data (open data) provided by the government. In the same vein, all posts published on Zlovenija were taken from publicly available (‘open’) profiles on Facebook, regarding which one cannot reasonably expect privacy.⁹ The method Zlovenija used nevertheless seems more severe, since it exposed ordinary individuals, not public figures (see Table 8.1).

On the benefits of open data for CDSTs

Open data has entered mainstream discussions, as the example of the Open Data Charter signed by the G8 (Group of Eight) leaders testifies. The Charter establishes several principles that all G8 members should observe in order to ‘help unlock the economic potential of open data, support innovation and provide greater accountability’ (G8, 2013). The benefits of open data have been well documented in the literature and include contentions that open data (1) leads to increased transparency and accountability with respect to public bodies and services; (2) increases the efficiency and productivity of agencies and enhances their governance; (3) promotes public participation in decision-making and social innovation; (4) fosters economic innovation as well as job and wealth creation (Kitchin, 2013); and (5) has an important internal value for the public sector itself as it gains access to data held by its other parts, or can make new use of its own structured data, improved by feedback from the public (Janssen, 2012, p. 4).

The analysed CDSTs to some extent reflect the benefits of open data. For instance, the creators of the web application Erar claim that the tool’s goals are to increase the level of responsibility of public office holders for effective and efficient use of public funds, to facilitate debate on adopted and planned investments and projects, as well as to decrease the risk of illicit management and abuse of office and to limit systemic corruption, unfair competition and clientage in public procurement procedures (Kovačič, 2016). Similar benefits can be established for the tools Who Influences?, Parlameter and the Legislative Activity Violation Counter as they all shed light on the operations of state officials and bring state power under increased scrutiny. Similarly, the immediate end of the whistle-blowing portal Žvižgač is to make those in power accountable for their actions.

However, such claims as to the benefits of CDSTs can hardly be made for the other analysed tools, i.e. iSee and Redaar. While iSee remains merely a tool for avoiding the surveillance gaze of the state and corporate sphere that does not directly contest the surveillance apparatus, Redaar goes even further and does not augment the accountability of the state, but undermines its effectiveness in pursuing the legitimate aim of penalising disorderly behaviour. Moreover, the CDST Zlovenija differs from the other analysed tools as it is not based on open government data (OGD) but on public Facebook posts. While it pursues a similar goal, which is to hold users accountable for their hateful comments about the migrants, it nevertheless is differentiated from the other CDSTs addressed in this chapter in one important aspect. The method of modern-day pillory that Zlovenija used was troublesome, and so were the targets it aimed to expose, as it was directed towards individuals for whom a public interest in the disclosure of personal data does not exist. Even Zlovenija’s authors agree that ‘it should not exist’, since it could have immense negative impacts on the lives of the exposed individuals, therefore raising both legitimacy as well as legality concerns. Although its authors ultimately decided that the goal was worth the risk, this played an important role when deciding to discontinue Zlovenija’s activities and ultimately led to its hibernation (Interview with Zlovenija, 2018).

Critiques of CDSTs

Kitchin (2014) outlines four critiques of open data: (1) it lacks a sustainable financial model; (2) it promotes a politics of the benign and empowers the empowered; (3) it lacks utility and usability; and (4) it facilitates the neoliberalisation and marketisation of public services.¹⁰ Let us examine these critiques in relation to the analysed CDSTs and add a few more.

The first argument, which refers to the sustainability of funding in the absence of direct governmental support, could be further developed by distinguishing between financing the initial publishing of open data and funding subsequent open-data projects enabled thereby. One step is to make data accessible, yet another is to extract meaning from data, the role assumed by infomediaries. The second step is particularly problematic, especially if the tools are developed by NGOs and engaged individuals that – in the absence of financial support from the government – are completely dependent on private funds, which is the case for the majority of the presented CDSTs in Slovenia (Erar being the only exception). The economics of creating sustainably funded initiatives is quite evident in the iSee and Redaar tools, which are no longer available to users. The Redaar application also shows how insufficient funding can lead to the end of the development of a particular tool. It has to be taken into account, however, that the aims Redaar attempted to pursue might be considered illegitimate, therefore resulting in significant obstacles to securing private funding in the first place. At the other end of the spectrum, Erar, the only state-funded CDST examined, is still up and running, albeit in a modified version due to power struggles over the balance between the public interest in access to information on public spending, on the one hand, and personal data protection, on the other.

In his second argument, Kitchin (2013) rejects the premise that everyone can access and use open data, leading to more engaged citizen participation and empowerment of society. On the contrary, he argues that this process is conditional upon: first, the means of processing large data sets (hardware and software); second, the knowledge and skills needed to analyse and interpret them; and finally, public support and the political will to enable policy change. Kitchin (2013) further noted that ‘even if some groups have the ability to make compelling sense of the data, they do not necessarily have the contacts needed to gain a public voice and influence a debate, or the political skill to take on a well-resourced and savvy opponent’. From all of the analysed CDSTs, the importance of the latter is particularly obvious in the case of Parlameter. Although its developers have both the means and skills to make sense of the information about the parliamentary process, there is little evidence that the tool actually had any impact on either political participation or oversight regarding the procedures of the National Assembly. In fact, Erar was the only analysed CDST that facilitated change on the national and regional levels in Slovenia. At the same time, it is also the only tool developed and financed by the government, which entirely supports Kitchin’s claim that open data empowers the empowered.

Open-data protagonists view data as neutral and not something embedded in the wider political, social and cultural contexts and contaminated with existing antagonisms. However, the open-data endeavour is never politically neutral or inherently a good thing. The politics of the data matters, and it is important to ask what the data reveal, how they are then used and in whose interests. In various contexts, the critical big data approaches have revealed how big data carries the allure of objectivity. For instance, in the predictive policing context, Joh (2017, p. 294) rightly observes that the police still have a creative role in creating inputs for automated calculations of future crime: ‘Their choices, priorities, and even omissions become the inputs algorithms use to forecast crime’. In his analysis of the politics of statistics, Desrosières (2002, p. 2) claims that even formalised synthetic concepts, such as averages, standard deviations, probability, identical categories or ‘equivalences’, correlation, regression, sampling and so on, are ‘the result of a historical gestation punctuated by hesitations, retranslations and conflicting interpretations’. Data are never ‘clean’, that is, devoid of social, cultural and economic circumstances.

It is then utopic to expect that open data will systematically and consistently over a longer period of time harm those in power. As our analysis of the Erar tool shows, the tool was able to cause political turmoil only in the short term. After the revelations regarding the risk of corruption in relation to the prime minister and one of his ministers, it ceased to include ‘dangerous’ data sets regarding the political elite and even the leadership of the CPC Commission was soon forced to resign due to political pressure.

Third, closely connected to this argument is the question of the utility and usability of the published data, since websites providing open data are often reduced to data dumps that instead of unveiling previously undisclosed information only provide a place to hide among the incomprehensible amount of published open data. Moreover, CDST developers in Slovenia are facing severe obstacles related to the quality and quantity of data. For example, Parlameter is based on data published by the National Assembly in a non-machine-readable format, which severely complicates CDST development. The challenges this presents are huge, but not insurmountable. On the other hand, Who Influences? is completely dependent on self-reported data, that is, the data reported to the CPC, which are incomplete or even entirely lacking with regard to some stakeholders.

Fourth, Kitchin (2014) claims that open data facilitates the neoliberalisation and marketisation of public services. The argument opens up the question of the context of data policies, that is, the question of the time and place of the implementation of open data. These determine how open-data policy is understood and how its meaning is attached and dependent on other social and political circumstances. While OGD started as a mainly UK and US driven initiative (Janssen, 2012), it has increasingly gathered attention in many other countries and international organisations, including the Organisation for Economic Co-Operation and Development (OECD), which had a pivotal role in ‘persuading’ other countries to adopt the initiatives. However, in other countries the idea of open data attaches itself to the specific political and societal processes thereof. The contextualisation of specific CDSTs in space and time provides insight into the implementation of an abstract notion of OGD.

In the Slovenian case, open data has had a specific meaning and is clearly attached to the transition process from socialist autocracy to capitalist democracy. OGD was even more enthusiastically embraced than elsewhere, as it has been taken up by the specific politics of transition with the aim of achieving specific political ends and signifies much more than open data actually can deliver. The argument underpinning the open-data paradigm in Slovenia has typically been as follows: due to the fact that in the autocratic socialist system government data were systematically concealed from the public and secrecy was the norm, now, under a democratic regime, the state must provide the data it gathers. By doing so the government will almost automatically become more democratic, more accountable and less tyrannical. Open data hence carries a liberation sprit of breaking with the ‘old and ugly’ autocratic system. Moreover, open data is coupled with the ideal of a market-oriented political system as it symbolises not only a turn towards democracy vis-à-vis autocracy, but also towards capitalism and away from socialism. Open data is supposed to make the newly developed state more capitalist. The state bureaucracy, the argument goes, was a burden for society and a parasite on the economic sector. The government must become lean, ‘vital’, more efficient and subject to market economy laws. Its assets, including data, must be monetised. In such a system, transparency is also a means of economic growth. In the post-socialist transition, open data then serves a specific political end. This is ‘the train’ (a metaphor particularly frequently used in political discourse) that society must embark on in order to escape its haunting past. In such a sociocultural context it is not common to find critiques of open data. Critiques are politically perceived as belonging to the demonised past, as open data is the means of transforming the country into a ‘normal’ democratic capitalist state.

Our analysis, nevertheless, does not suggest abandoning the move towards opening up data. We assert that open-data initiatives are political per se and that they attach to specific policies already existing in a particular time and space. Such initiatives flourish because they attach to other political and economic processes, which they in turn help amplify. Open data, as any other sociotechnical system, is inescapably political in its uses. By protecting, for example, environments characterised by social, political and economic inequalities, it can reinforce the conditions already at work there, to the detriment of social equality, justice and social cohesion. From such an understanding of open-data initiatives, we can observe how each of the analysed CDSTs requires that we be mindful of what data are being made open and how they are being funded, how data are made available, how they are being used and who is empowered by open data.

Turning to the analysed CDSTs, Supervizor is a pivotal example of how its power, which triggered a change in government, raised concerns due to destabilising the political status quo. However, Supervizor also provoked fear of ‘going too far’ with open data. It is not surprising, then, that in the years after the change in government in 2013, Supervizor was renamed Erar and the disputed data about budgetary payments in addition to individuals’ salaries – and that led the minister of education, science and sport to resign – were removed from the application. The privacy commissioner changed the interpretation of the conflicting values at stake: personal data protection now overrides the public interest in knowing how public funds are spent.

In our analysis we identified an additional critique of open data. Thus, fifth, open data may often infringe upon other competing values and step into an ethical borderland. The examined CDSTs reveal how the publication of some types of data sets poses a risk to personal data protection and a threat to individuals’ safety. The Zlovenija website went into hibernation relatively quickly, just 14 days after setting up the mentioned Tumblr page containing hateful Facebook comments, due to public pressure from the relevant public opinion makers (e.g. the eloquent former information commissioner) to protect the ‘culprits’ and not employ a vendetta or an ‘eye for an eye’ modus operandi. According to its authors,¹¹ they themselves were afraid for the life and limb of the exposed Facebook users. Moreover, the users themselves massively expressed apologies for their reckless trolling and requested that the Zlovenija moderator remove their posts on Tumblr with an explanation that the original Facebook posts had been removed.

Redaar and Zlovenija show how empowering CDSTs can quickly slide into an ethical grey area. On the one hand, Redaar was primarily used not to empower – in the sense of encompassing positive change, legal compliance or at least progressive social ends – but to breach the law by circumventing parking regulation enforcement. On the other hand, Zlovenija shows how the shaming of shamers can trigger clashes between human rights. The transparency of the shamers’ comments, either voluntary or not, may conflict with their security. While it is unconvincing to claim that social media users who publicly post hateful messages have a reasonable expectation of privacy, there are legitimate safety concerns when exposing them not only in online settings but also in the physical environment, as others have started to print out online comments and physically post them at various locations throughout the capital city of Slovenia, Ljubljana. Hateful comments exposed in different settings may trigger another wave of hatred, which goes in both directions: either targeted towards migrants, that is, the initial haters’ targets, or targeted towards the haters themselves.

Conclusion

Open data is often seen not just as a building block of citizen empowerment, but as the only prerequisite to achieving active participation, leading to a transparent and accountable governance regime. This optimistic view goes hand in hand with the big data ideology that we just have to collect enough data and ‘let them speak’ (Mayer-Schönberger and Cukier, 2013) as they will solve problems by themselves. However, this perspective is overly simplistic and entirely overlooks the fact that open data is not by itself neutral and is not used in a social vacuum, but will be weighed against the current political, economic and social background by the respective stakeholders, who may use it either as a tool to facilitate democracy, transparency and accountability, or to pursue their own contingent – possibly illegitimate or even illegal – agenda.

The work of the citizen data scientist is political. The data need to be processed, analysed and interpreted in order to give them their ‘true’ meaning – what that is will ultimately be decided by the CDST developers who have the power to present the data in a way that best suits their own perception of the world and the aims they are trying to pursue. It remains less clear how much power the political work of data scientist projects entails. While some of the projects funded and supported by the government, such as Erar and Supervizor before it, have gained significant political power in Slovenia, the remaining tools did not have comparable results, either in the form of political change or in the consequent engagement. Regarding the latter, we are learning that open-data tools in general do not activate otherwise passive citizens, but rather they provide a new tool for the already engaged. The question therefore remains, are CDSTs bridging or deepening the ‘data divide’?

As with any other technologically supported tool, CDSTs can be used either way. They could deepen the divide between the ‘big data poor’ and the ‘big data rich’ (Boyd and Crawford, 2012) and empower the elites to the detriment of the majority of the population, or they can genuinely serve as a catalyst of societal change. Moreover, when we look at the data, there is always a question of what we are looking at and what we are looking for. When you work with data there is always a story behind the processing of them. There always a question of normativity: what should and should not be there for us to find. Finally, rather than as the interpretations of information that they are, we too often see them simply as representations and descriptions of objective reality.

The authors are grateful to the reviewers for their comments. Special thanks are also owed to the funding Slovenian Research Agency, under research project ‘Automated Justice: Social, Ethical and Legal Implications’, no. J5-9347.

References

Andrejevic, M. (2005) ‘The work of watching one another: lateral surveillance, risk, and governance’, Surveillance & Society, 2 (4): 479–97.

— (2007) iSpy: Surveillance and Power in the Interactive Era (Lawrence: University Press of Kansas).

Ball, K.S. and F. Webster (eds.) (2003) The Intensification of Surveillance (London: Pluto Press).

Boyd, D. and K. Crawford (2012) ‘Critical questions for big data: provocations for a cultural, technological, and scholarly phenomenon’, Information, Communication, & Society, 15 (5): 662–79.

Centre for Information Service, Co-Operation and Development of NGOs (CNVOS) (n.d.) ‘Violations counter’, https://www.cnvos.si/nvo-vseved/stevec-krsitev/o-stevcu-krsitev (accessed 25 March 2020).

Commentality (2019) https://commentality.djnd.si (accessed 25 March 2020).

Compaine, B.M. (2001) The Digital Divide: Facing a Crisis or Creating a Myth? (Cambridge, MA: MIT Press).

Desrosières, A. (2002) The Politics of Large Numbers: A History of Statistical Reasoning (Cambridge, MA: Harvard University Press).

Eaves, D. (2013) ‘The importance of open data critiques – thoughts and context’, 18 Nov., https://eaves.ca/2013/11/18/the-importance-of-open-data-critiques-thoughts-and-context (accessed 25 March 2020).

Facebook (2018) ‘Facebook statement of rights and responsibilities’, 31 Jan., https://www.facebook.com/legal/terms/previous?ref=new_policy (accessed 25 March 2020).

G8 (2013) ‘G8 open data charter’, https://opendatacharter.net/g8-open-data-charter (accessed 25 March 2020).

Gigler, B.S. and S. Bailur (eds.) (2014) Closing the Feedback Loop: Can Technology Bridge the Accountability Gap? (Washington, DC: World Bank Group).

Gligorović, V. (2016) ‘Začetek družbenega eksperimenta’, Mladina, 7 Dec., https://www.mladina.si/177680/zacetek-druzbenega-eksperimenta (accessed 25 March 2020).

Granickas, K. (2013) Parliamentary Informatics: What Data Should Be Open and How Multi- Stakeholder Efforts Can Help Parliaments Achieve It, European Public Sector Information Platform, Topic Report No. 2013/05, https://www.europeandataportal.eu/sites/default/files/2013_parliamentary_informatics.pdf (accessed 25 March 2020).

Green, S. (1999) ‘A plague on the panopticon: surveillance and power in the global information society’, Information, Communication and Society, 2 (1): 26–44.

Institute for Applied Autonomy (IAA) (n.d.) ‘NYC surveillance camera project’, https://alternative-narratives-vis-archive.com/case_studies/isee.html (accessed 25 March 2020).

— (2004) ‘The influencers’, https://theinfluencers.org/en/institute-applied-autonomy (accessed 25 March 2020).

Integrity and Prevention of Corruption Act (2010) Official Gazette of Republic of Slovenia, no. 69/11, https://www.pisrs.si/Pis.web/pregledPredpisa?id=ZAKO6240 (accessed 25 March 2020).

Interview with Zlovenija (2018) Authors’ archive, Dec.

Janssen, K. (2012) ‘Open government data: right to information 2.0 or its rollback version?’ ICRI Research Paper, No. 8, 1 Sept., https://ssrn.com/abstract=2152566 (accessed 25 March 2020).

Joh, E.E. (2017) ‘Feeding the machine: policing, crime data, & algorithms’, William & Mary Bill of Rights Journal, 26 (2): 286–302.

Kitchin, R. (2013) ‘Four critiques of open data initiatives’, 27 Nov., https://blogs.lse.ac.uk/impactofsocialsciences/2013/11/27/four-critiques-of-open-data-initiatives (accessed 25 March 2020).

— (2014) The Data Revolution: Big Data, Open Data, Data Infrastructures and Their Consequences (London: Sage).

Koskela, H. (2006) ‘The other side of surveillance. Webcams, power and agency’, in D. Lyon (ed.), Theorizing Surveillance: The Panopticon and Beyond (Cullompton: Willan), pp. 163–81.

— (2009) ‘Hijacking surveillance? The new moral landscapes of amateur photographing’, in K. Franko Aas, H. Oppen Gundhus and H. Mork Lomell (eds.), Technologies of InSecurity: The Surveillance of Everyday Life (Abingdon: Routledge-Cavendish), pp. 147–67.

Kovačič, M. (2016) ‘Supervizor’, https://telefoncek.si/static/2016/06/Supervizor-IJS_2016_ENG.pdf (accessed 25 March 2020).

Malik, K. (2019) ‘As surveillance culture grows, can we even hope to escape its reach?’, The Guardian, 19 May, https://www.theguardian.com/commentisfree/2019/may/19/as-surveillance-culture-grows-can-we-even-hope-to-escape-its-reach?CMP=share_btn_tw (accessed 25 March 2020).

Mandelbaum, A.G. (2011) ‘Strengthening parliamentary accountability, citizen engagement and access to information: a global survey of parliamentary monitoring organizations’, National Democratic Institute, https://www.ndi.org/global-survey-parliamentary-monitoring-organizations (accessed 25 March 2020).

Mann, S., J. Nolan and B. Wellman (2002) ‘Sousveillance: inventing and using wearable computing devices for data collection in surveillance environments’, Surveillance & Society, 1 (3): 331–55.

Marx, G.T. (2002) ‘What is new about the “new surveillance”? Classifying for change and continuity’, Surveillance & Society, 1 (1): 9–29.

Mayer-Schönberger, V. and K. Cukier (2013) Big Data: A Revolution That Will Transform How We Live, Work, and Think (Boston, MA: Eamon Dolan/Houghton Mifflin Harcourt).

Mayer-Schönberger, V. and Z. Zappia (2011) ‘Participation and power: intermediaries of open data: conference draft’, draft paper prepared for the first Berlin Symposium on Internet and Society, 26–7 Oct., Berlin, https://www.hiig.de/wp-content/uploads/2017/08/participation_and_power.pdf (accessed 25 March 2020).

Monahan, T. (2006) ‘Counter-surveillance as political intervention?’ Social Semiotics, 16 (4): 515–34.

Monahan, T., D.J. Phillips and D. Murakami Wood (2010) ‘Surveillance and empowerment’, Surveillance & Society, 8 (2): 106–12.

Nakashima, H., H. Aghajan and J.C. Augusto (eds.) (2010) Handbook of Ambient Intelligence and Smart Environments (Boston, MA: Springer).

Norris P. (2001) Digital Divide: Civic Engagement, Information Poverty, and the Internet Worldwide (Cambridge: Cambridge University Press).

Oblak Črnič, T. (2017) ‘Fenomen ZLOvenija: v precepu med digitalnim aktivizmom in omrežnim individualizmom’, Časopis za kritiko znanosti, 45 (268): 18–198.

Plahuta, M. (2015) ‘Analysis of traffic violations in Slovenia between beginning of 2012 and end of 2014’, http://virostatiq.com/slovenija/prometni-prekrski-2012-2014/index-en.html (accessed 28 March 2020).

Rao, L. (2011) ‘Wael Ghonim: if you want to liberate the government, give them the internet’, TechCrunch, 11 Feb., https://techcrunch.com/2011/02/11/wael-ghonim-if-you-want-to-liberate-a-government-give-them-the-internet (accessed 28 March 2020).

Resolution on Legislative Regulation (2009) Official Gazette of Republic of Slovenia No. 95/09, http://www.pisrs.si/Pis.web/pregledPredpisa?id=ZAKO5516 (accessed 25 March 2020).

Stallwood, O. (2013) ‘Game to destroy CCTV cameras: vandalism or valid protest?’, The Guardian, 25 Jan., https://www.theguardian.com/theguardian/shortcuts/2013/jan/25/game-destroy-cctv-cameras-berlin (accessed 25 March 2020).

Trottier, D. (2017) ‘Digital vigilantism as weaponisation of visibility’, Philosophy & Technology, 30 (1): 5–72.

van Dijk, J.A.G.M. (2006) ‘Digital divide research, achievements and shortcomings’, Poetics, 34: 221–35.

Virostatiq (2015) ‘Analysis of traffic violations in Slovenia between beginning of 2012 and end of 2014’, http://virostatiq.com/category/mobile-apps (accessed 25 March 2020).

Warschauer, M. (2004) Technology and Social Inclusion: Rethinking the Digital Divide (Cambridge, MA: MIT Press).

Završnik, A. (2014) ‘Videonadzor v. Sloveniji’, Pravna praksa, 33 (35): 33.

Završnik, A. and P. Križnar (2018) ‘Legal standards of location privacy in light of the mosaic theory’, in B.C. Newell, T. Timan and B.J. Koops (eds.), Surveillance, Privacy and Public Space (Abingdon: Routledge), pp. 199–220).

Zlovenija (2015) http://zlovenija.tumblr.com (accessed 25 March 2020).

Zuboff, S. (2015) ‘Big other: surveillance capitalism and the prospects of an information civilization’, Journal of Information Technology, 30 (1): 75–89.

Žvižgač (2019) https://zvizgac.si (accessed 25 March 2020).