The Promise of the Transparency Culture

Abstract

In 2013, with the introduction of the Transparency Law in Spain, a number of Spanish newsrooms started working with data journalism methods. In Sweden, which has one of the oldest Freedom of Information acts in the world, newsrooms invested in the skill development of data journalism at approximately the same time. Because previous research suggests that access to public data has been one of the key driving forces for the development of data journalism worldwide, it is important to understand how legislation is actually shaping the practice of data journalism. Based on a survey of 66 key informants in Spain and Sweden and ten in-depth interviews with data journalists from five media companies in each country, we conducted a comparative study, building on the frameworks of media systems to explore data journalism practices in these two countries. The differences found indicate that the national and EU legislation in both nations shape journalistic strategies for accessing data, turning journalists at times into activists fighting for the right to access public data. Beyond the law, data journalists advocate for a transparency culture among the civil servants, in order to secure public accountability.

KEYWORDS:

• data journalism
• freedom of information act
• media systems
• public data
• transparency

Introduction

The global rise of data journalism in recent years has equipped newsrooms all over the world with journalists able to process, analyse and present stories based on large amounts of data (Borges Rey 2017; Uskali and Kuutti 2015). In this context, data is information, often unstructured, collected through Freedom of Information requests, by scraping websites, through leaks, by data feed subscriptions, data licenses (Borges Rey 2017), and with the aid of crowdsourcing techniques (Appelgren and Nygren 2014). Potentially, these datasets can be used to uncover stories that are otherwise outside the public domain (Cushion, Lewis, and Callaghan 2017), which is one of the normative prerequisites for investigative journalism. This a form of journalistic interventionism (Reunanen and Koljonen 2018, 735) that requires a systemic precondition: the transparency of public institutions when providing data. For data journalists, public records often play a central role in the creation of data journalistic projects (see, for example, Young, Hermida, and Fulda 2018). Even though there is a substantial number of studies on the ongoing development of data journalism (Ausserhofer et al. 2017), little focus has been paid to the sometimes cumbersome and extremely time-consuming task of asking authorities for access to what is paradoxically often referred to as open data or public data.

Cross-national comparative studies allow the exploration of nuances of socio-professional phenomena. Comparing Spain and Sweden seems particularly suitable for this purpose: both are democratic societies with well-developed media systems, but they show significant differences in their respective approaches towards public transparency (Salaverría and Gomez Baceiredo 2017; Wadbring and Ohlsson 2017). In Sweden, the principle of free access to public records has historically allowed all citizens to access documents handled by the authorities, while citizens of many countries around the world, including Spain, have only recently granted such a right.

In 2017, more than half of the EU countries had laws guaranteeing open decision-making, citizen participation, and overall transparency in public governance (MHE 2017). Since the Treaty of Lisbon entered into force in 2009, the EU itself has fostered transparency within its institutions (Curtin and Leino-Sandberg 2016). Nevertheless, it is still difficult for journalists to access public data, and in this paper, we aim to explore the strategies currently used by data journalists in two countries with a distinct history of legislation and openness to public data. The main research questions are:

RQ1: What strategies are Spanish and Swedish data journalists currently using to obtain public data?

RQ2: What are the major challenges facing the development of data journalism in Spain and Sweden in relation to national legislation stipulating freedom of expression and procedures related to access to public data?

Background to the Transparency Laws

The Swedish Freedom of Press Act, which sets forth access to public information as a citizen's right (Luthanen 2006), was passed in 1766 and is considered to be the oldest Freedom of Information legislation in the world (Riksdagen 2016). In the EU, the first regulation on access to public documents was written in 2001, and although many EU countries have already passed transparency laws, the performance of these laws and the general public's perception of public transparency vary notably. Another important transparency initiative in Europe was the PSI (Public Sector Information) Directive adopted in November 2003 by the European Commission (Directive 2003/98/EC) (Aitamurto, Sirkkunen, and Lehtonen 2011). This directive encouraged re-use of public sector data to make it easier for third parties to create new services, but it also provided journalists with easier access to digital public data.

In terms media system development, Hallin and Mancini (2004) found Spain to belong to the Polarized Pluralist Model, while Sweden was considered to be one of the classic examples of the Democratic Corporatist Model (Hallin and Mancini 2004). Albarran (2009) claims that the Spanish media has transitioned to the Democratic Corporatist Model, aided by technology and a liberalization of the media from government control. Ohlsson (2015) argues that there is an ongoing European trend, including in Sweden, of media systems converging towards the Liberal model, which is characterized by deregulation and commercialization. Despite this homogenization, there are some distinctive features that remain in the Spanish media system: low circulation of newspapers, pronounced tendencies towards commentary, weak professionalization of journalists, and strong state intervention (Salaverría and Gómez Baceiredo 2017). Sweden's media system is still one of the strongest newspaper markets, with a high degree of professionalization and a declining, yet still strong, state role in media policy-making (Wadbring and Ohlsson 2017). According to Svensson and Edström (2016), Sweden is built on a democracy-driven rationality of freedom of expression. The state is required to intervene to uphold free speech, and one of its responsibilities is to uphold free media, in that public authorities are not permitted to examine in advance what is broadcast or printed (Riksdagen 2016).

Among the large and highly populated European countries, Spain was long the only country lacking a Freedom of Information act. Although the right to access public information was formulated as a law in 1992, it was not until 2013 that the Spanish Transparency Law was finally passed (La-Rosa and Sandoval-Martín 2016). As in other similar legislation, this law has three main objectives: to guarantee the right to access information for all citizens, to enforce transparency of all public administrations, and to supervise the observation of fair governance practices. For the purpose of the third objective, a Portal of Transparency (http://transparencia.gob.es/) was opened in December 2014 by the central government, and many regional and local authorities followed suit. However, the law is subject to criticism concerning access to information does not yet comply with the European Court of Human Right's jurisprudence (Romanos 2017). There are also concerns about the difficulty of accessing information as well as the obstacles to obtain data about public contracts (Beltrán-Orenes and Martínez-Pastor 2016).

Transparency and Journalism

Classic theories of journalism have, for decades, recommended providing as much information as possible as one of the principles for the reliability of news, not only with regard to sources but also with regard to files and archives, as a way of reinforcing objectivity. Previous advocates of a new data-driven journalism (Meyer 1991), besides requiring details about sources and content used to research news, began to also demand full disclosure of the methods journalists used to gather and analyse information. This led journalism theorists to shift from the classic normative demands of objectivity, an increasingly questioned concept since the 1980s (Hackett 1984; Schudson 1981), towards a new claim for “transparency” (Karlsson 2010) as the “new objectivity” (Vos and Moore 2018, 12).

Today, the open nature of how data can be accessed and presented in the digital environment of the internet results in the fact that journalists expect transparency from those they cover, i.e. key institutions of social power—government, business, religion, and health care (Vos and Craft 2017). However, the practice of journalists accessing open data is far from the normative ideas of fostering democratic processes by promoting transparency through the publication of government datasets (Ruijer, Grimmelikhuijsen, and Meijer 2017, 45), as journalists most often have to face the bureaucratic mechanisms of public bodies with lengthy negotiations that, in the worst case, may result in being provided data in a non–machine-readable format (Borges Rey 2017).

Transparency has expanded in today's practice of journalism, which occurs in multiple spheres of the media (Craft and Heim 2009). News media companies are increasingly urged to provide full disclosure about their ownership conflicts and business interests (Baker 2006) and the need for transparency has been extended to reporters: there is a growing demand of accountability in the way information is accessed and handled (Phillips 2010). Finally, there is an increasing requirement for full transparency and open access to sources of information, especially in public institutions (Bertot, Jaeger, and Grimes 2010).

This multiplicity of spheres affecting transparency has produced studies on the ways the requirement of greater transparency has influenced professional practices in the newsrooms (See for example Chadha and Koliska 2015). These changes seem to be transforming the traditional “rituals of objectivity” (Tuchman 1972), widely observed by journalists for decades, into new “rituals of transparency” (Karlsson 2010), in which there are differences in the “scale” and the “techniques” (Karlsson 2010, 542). We can hypothesize, then, that the appropriation of this ritual among journalists has expanded over time and that the way this concept is interpreted and implemented today differs not only between media outlets but also—and especially—between countries and media systems.

Methods

For this study, we used semi-structured interviews with ten key informants working at five Spanish and five Swedish news organizations. Furthermore, we based our findings on a survey directed to 96 Spanish and 84 Swedish data journalists that contained 17 questions about data journalism and public data.

In Spain, we interviewed journalists from the newspapers El Mundo and Diario de Navarra, digital natives El Confidencial and Nació Digital, and a news start-up, Datadista. In Sweden, we interviewed data journalists from Swedish Public Television (SVT), Swedish Public Radio (SR), two news start-ups, J++ and United Robots, and the national trade press, Lärarnas Tidning.

We then set up a database of data journalists that consisted of a non-probability purposive sample of 96 identified Spanish data journalists and 84 Swedish data journalists. Spanish journalists were invited to participate in the online survey by sending them a personal message on Twitter with a link. In Sweden, because Twitter is not as popular as in Spain for news (Newman et al. 2017, 11), invitations were sent via email. Response rates were 46% in Spain and 25% in Sweden. The choice of two different contact platforms may be the reason for the lower response rate among the Swedish journalists, as Twitter is a more immediate and less-formal medium for making contact compared to e-mail.

The Spanish semi-structured interviews took place 7–23 February 2017, in Madrid, Barcelona and Pamplona at places chosen by the subjects, such as a conference room at their news organization or in a nearby café. The Swedish interviews took place in Stockholm and on Skype 18–31 August 2017. The interviews were between 40 and 50 minutes long.

Results

The online survey results show both similarities and differences in how data journalists perceive the process of asking for public data compared to other challenges that data journalists are currently facing in the two countries. Journalists engaging in FOIA requests indicate that they have planned to create stories that other media outlets may not have. This is in line with how Reunanen and Koljonen describes journalistic interventionism (2018, 735). The majority of the Swedish journalists (71.4 per cent) state that they have asked for data from public authorities a few times, compared to 50 per cent for the Spanish journalists, and almost one-fifth of the Spanish data journalists have never asked for public data (Table 1). Swedish data journalists thus seem to be more prone to journalistic interventionism (Reunanen and Koljonen 2018).

TABLE 1 Perceived frequency of asking for public data, per cent (n = 45, 21)

Have you ever asked for data (big/raw data) from a public authority?	Spain	Sweden
Yes, I do it all the time	23.8	14.3
Yes, a few times	50	71.4
Yes, once	7.1	4.8
No	19	2

The recent introduction of the Spanish transparency legislation may, however, be the main reason for the fact that access to public data is still limited among Spanish data journalists. More than half of the Spanish journalists only somewhat agree that authorities provide digital data, compared to one-third of the Swedish journalists. Table 2 shows attitudes towards the journalistic interest in published public data. As previous research has found that the open data movement has shaped the practice of data journalism in Sweden (Appelgren and Nygren 2014), Swedish data journalists may be more accustomed to the kinds of public data that authorities publish. Spanish journalists express a more negative attitude towards the nature of data published, while more than one-third of the Swedish journalists in our sample agree that the data published is of journalistic interest. These results may, however, also be linked to fact that the practice of making FOIA requests seems to be more common in Sweden than in Spain.

TABLE 2 Attitudes towards the value of published public data, per cent (n = 45, 21)

The data that public authorities publish are not interesting	Spain	Sweden
Completely agree	9.5	0
Agree	33.3	20
Neutral	42.9	20
Somewhat agree	14.3	25
Do not agree	0	35

The introduction of the transparency portal in Spain occurred when the new law was introduced to help citizens access public data. It is the national gateway to public data in Spain. Journalists submit requests, which are then either granted or denied. There are also regional and local transparency portals, which can make the process more complex, but there is an advantage to having everything distributed in one place. Swedish journalists, on the other hand, must ask each individual authority for data, sometimes using portals from which data is accessed with the aid of APIs, but most often by requesting data from the staff working at the authorities. Due to the fact that the legislation was created before today's digital media environment, requests may often be granted as huge amounts of information printed on paper. Table 3 shows how data journalists value various challenges to data journalism. The Spanish journalists’ top challenges include: access to data, statistical skills, and public transparency. Simplifying the process of accessing public data does not seem to meet Spanish data journalists’ expectations in terms of the normative ideals of transparency in the digital environment. Furthermore, even though Swedish data journalists do not have a single digital transparency portal, they did not report that access to data was as challenging as it was for the Spanish data journalists. Instead, they reported support from company management, cooperation with non-journalists, and newsroom organization as the most challenging issues for the ongoing development of data journalism.

TABLE 3 Challenges to the development of data journalism, (Averages of 1 to 5, where 1 is No challenge and 5 is Major challenge) (n = 45, 21)

In your opinion, what is currently the biggest challenge to data journalism evolving in your country?	Spain	Sweden
Access to data	4	3.5
Access to the right technology	2.6	2.9
Computer skills (programming, user experience, design, etc.)	3.8	3.8
Cooperation with non-journalists	3.3	4.1
Legislation	3.3	3.2
Market-related issues	3.1	3.0
Newsroom organization	3.7	3.9
Professional training courses	2.8	3.5
Public transparency	3.9	3.2
Statistical skills (math, data analysis, etc.)	4	3.7
Support from company management	3.7	4.1
University education	3.7	3.5

The Swedish journalists’ perception of the challenges echoes what Borges Rey (2017) denotes as a techie approach, emphasizing collaboration between journalists and data professionals.

In-depth Interviews

The Spanish journalists, despite expressing some form of gratitude for their national legislation, stated dissatisfaction with the current performance of the system to access public information.

There are a lot of bad things about the law, but we want to view it from the bright side. We wouldn't have [access to public information] without the law. The main strength is that [the Transparency Law] exists [..] and that we have the Council of Transparency, which is totally independent.

Similarly, another Spanish journalist criticized openly how the Transparency Act has been implemented because it is secondary to other laws.

The law is not good enough. [..]The problem is that the Transparency Act may collide with other laws and these laws can take precedence over the Transparency Act.

The Swedish interviewees rarely reflected over the presence of the legislation; they took it somewhat for granted.

In Sweden, we are lucky to have very supportive national legislation, which makes us less prone to check for example EU legislation. [..] We are too comfortable, even though [other types of legislation] would make it possible for us to access more interesting data.

However, they did note that few of their colleagues take the effort to check what kind of public documents they can legally access.

Swedish journalists should be familiar with two laws, and there should be no doubt about their content: The Freedom of Press Act and the Public Access to Information and Secrecy Act. That is all you need.

With the support of these two laws, journalists can access, for example, documents within an authority describing the existence of all public documents at the authority and case processes handled within the authority. These documents speed up the process of finding the type of data that journalists need for a particular story. Yet, according to the interviews, many Swedish journalists are reluctant to make the effort to ask for them and, even more troubling, perhaps are also unaware that such documents exist.

Like the Spanish journalists, the Swedish informants also expressed dissatisfaction with the current organization or system to access public documents.

One thing [the public authorities] do is to deny you access to the information you have requested. Another thing they do is make it as difficult as possible for you to use the information you have requested: only give you the data printed on paper or deliberately update their data routines [when you have asked for data] to make it more difficult for you to access what you want.

Furthermore, in the Swedish case, both journalists and authorities seem to have found strategies to bypass the legislation, such as when information is considered classified.

I believe there are dilemmas with the secrecy examination at authorities. It happens quite often that someone have decided that certain information is classified. But, when I ask around and try to access the information from different people within the authority, I might be able to get hold of the information anyway.

The Spanish interviews did not indicate any tendencies to bypass the system in this manner, perhaps in part because the Spanish case is less complex. The Spanish informants’ complaints were mainly related to the denial of data and the amount of time it takes to receive approved information.

Asking for information through the portal is very slow, which we can't afford. If they do not want to give you the data, they say No right away. [Otherwise], you have to wait one to two months.

When time is not a crucial factor, journalists can create datasets over longer time periods. A few years ago, one of the Swedish informants wanted to access log files of lobbyists visiting MPs. She asked for all visits over a period of one week and received an Excel file containing the names of visitors, the MP they visited, and the duration of each visit. However, when asking for the very same data for every coming week for the remainder of the year, the request was denied. Then, as a precaution, the Parliament administration decided to alter their visitor log policy.

After my initial request, they decided to delete their log files every night, instead of every week. I had to submit a request every day, but legally I was not allowed to place a standing request. I had to submit a request by email every day, which I think is bullshit, to be honest. They also decided that I couldn't have the data in digital format; instead, I got it on paper. Each week, I got a bunch of paper with a printed copy of the digital visitor log files.

The experienced journalist then got help to build a case convincing the authorities that that the log files should be accessible as part of the Freedom of Information Act. They did not succeed, but their request had to be decided by the Prime Minister himself due to his status as the Head of State. This example and other comments during the Swedish interviews demonstrate that journalists are almost turned into activists. In general, Spanish journalists do not see themselves as activists, although they sometimes complain publicly, due to the lack of transparency in the public authorities. The Spanish interviews instead highlighted the importance of being clever in advance when formulating requests to the Transparency Portal.

Submitting an information request [to public authorities] is like playing Battleship: you have to tell them A2, C2 … and you have to sink the boat. You must know the law inside and out in order to know which questions will be rejected by the officials.

You learn how to write a question, to leave it not too open, but quite closed.

Central in all our interviews, asking for public data is attributed to a certain level of experience and competence; the interviewed journalists in both countries have developed often-complex strategies for accessing public records. In Sweden, they engage in seemingly harmless activism by attempting to “break the system” or not giving in when denied requests. In Spain, the journalists focus on how to pose undeniable questions to the Transparency Portal.

Discussion and Conclusions

The effects from the several-hundred-year difference between when the Freedom of Information Acts were first introduced in the two countries would require a much deeper analysis, but we found indications that there are distinct strategies adapted to the legislation and the accompanying system.

The results from the qualitative interviews may serve as an example of how an older Freedom of Information legislation, such as in Sweden, may influence journalists both positively and negatively, compared to how journalists in countries with newly introduced laws, such as in Spain, similarly perceive their situation as both positive and negative in relation to the legislation and the system that follows with it. In Spain, the main strategies were to pose a request to the Transparency Portal that could not be denied and to accept that data requests would take several months to process. The Swedish journalists, having requested data from authorities for a longer period time, emphasized that the knowledge and experience in asking for data presented solutions that would allow them to bypass the system when denied data.

In both countries, authorities are supposed to provide data, but the results point to a strong reluctance to give journalists the data they request unless forced to do so following legislative reminders. Activism, in terms of applying different kinds of pressure to access data when denied, was present in the Swedish interviews, but it was not clearly discerned in the Spanish interviews. The relative detachment from activism observed among the Spanish data journalists could be interpreted in line with Albarran's (2009) observation that the Spanish media system is evolving towards the Democratic Corporatist Model (Hallin and Mancini 2004) in which professionalism is central. Spanish journalists show notable disbelief in the power of the legislation to induce public authorities to be fully transparent, indicating that Spanish journalists think they are still working in an underdeveloped democratic media system.

Luthanen (2006, 57) argues that “the transparency of government is a core issue of the democracy principle and a precondition for bringing any government closer to its citizens”. However, we have found that the law is not enough to secure public transparency. Besides legislation, a society must share a culture of accountability, observed both by the citizens and their public servants. Such a distinct consideration of transparency as a civic value is what really seems to separate the Nordic and Mediterranean countries, but even with one of the oldest freedom of information acts in the world, legislation does not seem to guarantee public access to public documents.

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by Spanish Ministry of Economy and Competitiveness of the Government of Spain: [Grant Number CSO2015-64662-C4-1].

Notes on contributors

Ester Appelgren

http://www.sh.se/ester_appelgren; http://orcid.org/0000-0003-1428-9477

Ramón Salaverría

http://www.unav.es/cv/rsalaver/en/; http://orcid.org/0000-0002-4188-7811