The politics of internet freedom rankings

Section 1. Introduction

Today, internet freedom remains in the eyes of the beholder – or, in this case, the eyes of those who design the various indexes assessing it. International rankings and indexes play an active role in defining the issue they claim to capture (e.g. ‘development’): “By naming an issue, coining a vocabulary for describing it, and creating categories for its assessment, promulgators hope to affect discourse and ultimately policy” (Kelley & Simmons, 2019, p. 495). The creation of new international rankings, moreover, serves to give the issue salience; i.e. to present it as a matter of global concern. As internet access expanded globally, the past two decades have seen a rapid proliferation of rankings, indexes and dashboards measuring and comparing the state of internet freedom around the globe. This set of rankings – from Freedom House’s Freedom on the Net (first published in 2009) to Ranking Digital Rights’ Corporate Accountability Index (launched in 2015) – has, at times literally, helped put the issue of internet freedom on the map.

But why are there multiple, competing rankings? What can the development of this rankings landscape tell us about the shifting understandings of what internet freedom is, and about who stands to gain from how it is operationalised? As a cluster of communicative technologies and practices used to both oppose and support institutionalised power structures, the internet continues to be, paradoxically, at once a globalising force challenging the hegemony of nation-state governance models and a coercive tool for drawing new sovereign boundaries in this networked space (Kohl, 2017). Internet freedom rankings, like the popular democracy and media freedom measurement instruments before them, are powerful “global pattern-setters” for how internet freedom is understood, as rankings are often “considered as a condition (i.e. a sort of natural order of things) rather than a narration (i.e. a product of human action)” (Giannone, 2010, p. 70).

Instead of scrutinising an individual ranking, our aim is to understand how and why such a complex landscape of internet freedom rankings has emerged. In our approach, we draw inspiration from the relational approach to international rankings (Beaumont & Towns, 2021), which stresses the role of, and power relations among, relevant actors involved in the creation, promotion and responses to rankings. Analysing the development of this landscape, we identify how the ranking organisations’ varying approaches to capturing internet freedom have played a role in defining it and legitimating it as an issue of importance to be tended to. Acknowledging how indexes also serve “as a strategic tool for producing authoritative expertise – or at least the public appearance of expertise” for “NGOs and some [International Organisations], which frequently find themselves in competition with their peers for allies, attention, and resources” (Broome & Quirk, 2015, p. 824), we examine which ranking producers have managed to successfully establish themselves (and thereby their conceptualisation of internet freedom) and which ranking producers did not manage to do so. We approach internet freedom rankings as “public performances” that are carefully orchestrated by their respective ranking organisations in the public domain (Ringel, 2021a), rather than as mere measures of quantification and classification in need of independent validation.

The article is structured as follows. We first review the scholarly debate on international rankings. We then trace the evolution of internet freedom as a concept and the tensions inherent in its evolving definition. After introducing our methodological approach, we examine internet freedom rankings from an organisational perspective and as “public performances” (Ringel 2021a; 2021b). We interpret our findings against the development of (the understanding of) the internet itself and its societal embeddings, including shifting definitions of internet freedom. We provide vignettes of three rankings – Freedom House’s Freedom on the Net; Reporters without Borders’ Enemies of the Internet; Ranking Digital Rights’ Corporate Accountability Index – to illustrate the key trends and explain the diversity within the internet freedom rankings landscape.1 We find that the examined rankings, in many respects, perform similarly to rankings in other fields, e.g. higher education. Yet, we argue, the complexity of internet freedom – where both the uses of the internet and discussions about defining what freedom means in relation to it have developed rapidly – poses unique challenges and requires ranking organisations to be attuned to these developments and negotiate them with other actors in their field.

Section 2. The politics of international rankings

International rankings, also referred to as Country Performance Indicators (CPIs), Global Performance Assessments (GPAs) or Indicators (GPIs), are a comparative tool that serves as an evaluative shorthand in decision-making contexts. They are a technology of knowledge production, whose power lies in “their capacity to convert complicated contextually variable phenomena into unambiguous, clear, and impersonal measures,” as well as a tool of governance (Merry, 2011, p. S84). The 1990s and 2000s have seen a “nearly exponential” increase in the number of international rankings (Kelley & Simmons, 2019, p. 493), which, according to Cooley, resulted from the “adoption of techniques of performance evaluation in modern political and social life” inspired by neoliberalism; “the strengthening of global governance networks” in which rankings can also act as monitors to assess compliance with particular international standards or norms; and “the proliferation of new information technologies and open data sources” (Cooley, 2015, p. 10). The trend also coincided with the end of the Cold War and the assumed processes of democratisation, e.g. in Central and Eastern Europe, where rankings played an important role in monitoring progress. 

As they are created for a purpose and with particular audiences and decision-making processes in mind, rankings are necessarily political and relational. Regardless of the nature and affiliation of the actors involved in their creation, “rankings are designed to exert normative pressures on states to promote change in a country’s performance or improve some aspect of its domestic institutions or policymaking” (Cooley, 2015, p. 2). The capacity of rankings to induce changes in a state’s behaviour is assumed to follow from rationalist considerations of the material costs involved, e.g. affecting the allocation of foreign aid, or from how a ranking affects states’ international status (Kelley & Simmons, 2019). The latter works most effectively when “an international ranking highlights [the states’] hierarchical standing, either through ‘naming and shaming’ or by judging them against a peer state, rival, or regional grouping” (Cooley, 2015, p. 6). Here, rankings may also impede justified criticism aimed at states who score well on key indicators. Civil society organisations can draw upon global rankings to bolster their efforts and push for (domestic) policy change, or they can create their own indexes to frame and grant salience to a policy issue (Urueña, 2018). On the flip side, the use of rankings in (inter)national policymaking may determine whether civil society actors are allocated resources and how the impact of their activities is evaluated (Merry, 2011).

The methodology applied to measure and compare the phenomenon the ranking aims to capture can reflect the “political and ideological climate in which it was conceived” (Giannone, 2010, p. 70). Freedom House’s Freedom in the World index, for example, reflects a neoliberal understanding of democracy in which the value of liberty is key and civil and political rights take precedence over socio-economic rights (Giannone, 2010, p. 78). Such “global benchmarking” tools then, “serve to both ‘neutralise’ and ‘universalise’ [...] normative values and agendas” (Broome & Quirk, 2015, p. 819), often pertaining to the “global North” (Merry, 2011). Through their standard-setting power, the assumptions and values embedded in how a ranking operationalises its subject can become ingrained in “international institutions and administrative practices” (Cooley, 2015, p. 2).

Leading international democracy and media freedom indexes, e.g. those produced by Freedom House and Reporters without Borders, have been criticised regarding their methodological soundness, conceptual validity and ideological orientation (Broome & Quirk, 2015; Brooten, 2013; Giannone, 2010; ibid., 2014; Gunitsky, 2015; Landman, 2018). For example, global rankings “inevitably rely on local data-collection processes, although they may be created and managed at the international level,” which may lead to divergent understandings of, for example, how measurement tasks are performed (Merry, 2011, p. S89). Indeed, as Beaumont and Towns (2021, p. 1469) point out, the paradox of global rankings is that they “persist despite their often questionable nature” in terms of the “dubiousness of much data” used to compile them and their “problematic side effects,” such as states seeking to “game” the ranking. This observation underscores the importance of looking beyond the methodological aspects of a single ranking and, instead, considering how ranking producers and other relevant actors relate to each other, in order to understand how internet freedom has been defined and legitimated as an issue of global concern through rankings.

The extensive media coverage many rankings receive plays an enabling role in inducing policy change through social pressure but also reinforces the authoritative status of the rankings themselves (Beaumont & Towns, 2021). Rankings “become public measures through an array of organisational practices [...] aiming to raise the audience’s attention and to provide a rich experience” (Ringel, 2021a, p. 56). Publication is a carefully orchestrated “public performance” in which the definition of themes and narratives, the visualisation of the ranking, the selection of the launch date and format of the launch event itself are all tailored towards optimising (global) attention and engagement with target audiences (Ringel, 2021a). The smaller the change in the ranking table, the more important the accompanying narration becomes to still generate newsworthiness. Through press releases and (pre-publication) briefings, ranking organisations seek to steer media coverage to highlighted findings. As rankings, by the very fact that they create “winners” and “losers” are likely to draw critique, ranking organisations continuously work towards maintaining their credibility once they have successfully established their authority, e.g. through transparency (disclosing methodology, sharing data) and inclusivity efforts (proactively engaging [critical] stakeholders) (Ringel, 2021b). Providing extensive contextual materials aims to preempt the critique that rankings present an oversimplification of affairs.

Many of our everyday interactions with data, including that underlying internet freedom rankings (Bandola-Gill et al., 2021), are facilitated by forms of visualisation. They influence how we interpret data in ways that extend beyond the information contained in the visualisation itself and can “mitigate [political] framing effects” (Baumer et al., 2018, p. 22). Data visualisations, such as the presentation of rankings in (interactive) maps and charts, can “contribute to the formation of public opinion about contested matters” (Nærland, 2020, p. 65). Visualisations can “privilege certain views of the world” or act as “carriers of ideology” (p. 66), but can also be “used instrumentally to guide policy or decision-making” and “foster [citizens’] engagement with [political] processes and political participation” (Nærland, 2020, pp. 68-70). The standard format for presenting rankings – the league table – is poorly equipped to capture phenomena that are characterised by multiplicity (such as, we would argue, internet freedom), “as its main rhetorical focus is on the clarity of communication [...] of the ranking and in the immediate visibility of performance” (Bandola-Gill et al., 2021, p. 46). The trend towards presenting data in dashboards and other interactive visualisation formats is thought to “mitigate the competitive and potentially dysfunctional pressures of the display of ‘winners and losers’” (Bandola-Gill et al., 2021, p. 27).

To examine how internet freedom has been defined and legitimated as an issue of global concern through rankings thus requires an integrated and relational approach which comprises the rankings’ organisational background, methodology and production practices, and narration and visualisation through multi-modal communication practices. Charting the development of the landscape of internet freedom rankings over the course of two decades allows us to reflect on the differences in the strategies adopted by these competing rankings and on who stands to gain from the particular ways in which they have framed and operationalised internet freedom.

Section 3. A brief history of defining internet freedom

Early ideas of internet freedom have hinged on foundational concepts such as human rights and freedom of expression. They have also been shaped by various ideological drifts, from libertarian ideas about free will and individual freedom, exemplified by John Perry Barlow’s “Declaration for the Independence of Cyberspace” (1996), to meritocratic views of freedom and access for those most able and technologically literate. As the global internet infrastructure became more sophisticated, and the networks underpinning it more complex, the understanding of what constitutes internet freedom also evolved. Freedom of information and expression were framed as component parts of internet freedom, while more equitable control over the means of communication was highlighted as a counterbalance to hegemonic information flows (Hamelink & Hoffman, 2008). The McBride report, presented to UNESCO in 1983, summarised the push by countries in the “Global South” to recognise the “right to communicate” as a complementary notion to freedom of expression that guaranteed the rights of “marginalised peoples to not only receive information but express themselves and shape their own stories and images” (Brooten, 2013).

Isaiah Berlin’s (1969) early conceptualisation of liberty as two-fold – i.e. understanding internet freedom as either “freedom from” (negative liberty) or “freedom to” (positive liberty) – provides insights into the tensions inherent in assessments of internet freedom. The two notions of freedom are not understood as mutually exclusive, yet some policymakers insist that positive freedom aims to “protect and promote the rights of the public as a whole,” whereas negative freedom seeks to “protect and promote the rights of all individuals” (Ross, 2010). Both understandings have been applied by scholars, policymakers and digital rights activists in their attempts to articulate the contested nature of internet freedom.

 Along with aspects of positive and negative freedom, baseline human rights set forth by the Universal Declaration of Human Rights (United Nations General Assembly, 1948) generally provide the skeleton for such evaluations. Yet the cultural and economic diversity of the various national contexts complicates how these abstract principles are perceived. On the one hand, scholars have argued that it is possible to design multi-level models that demonstrate the complex relationship between the level of democratisation, internet use and internet penetration (Nisbet et al., 2012) although there may be discrepancies between perceived and actual levels of internet freedom (Stoycheff, 2020). On the other hand, opinion research involving internet users (Internet Society, 2012; Pew Research Center, 2014) suggests that demand for internet freedom is itself a function of internet use and digital literacy. The type of online activities may also mediate the effect of internet use in authoritarian states, where recreational use was found to be “associated with satisfactory evaluations of non-democratic regimes and more entrenched authoritarian worldviews” (Stoycheff et al., 2016, p. 1034). 

The correlation of regime type with levels of internet freedom has often been extrapolated from the relationship between media freedom and the level of democratic consolidation in a society. However, internet freedom is a far more contested concept than press freedom due to the global reach of its policy implications (Shen, 2017). In particular, freedom of access, for example to facilitate the delivery of governmental services and distance education, is more widely recognised by states than the freedom of use – to express opinions, share and receive information online – even though the latter is often assumed under the notion of internet freedom measured by international rankings (as we will demonstrate below). Researchers such as DeNardis (2014) have also documented the emerging tensions between the globalisation of internet governance and the rising concern with internet sovereignty among nation-states that transcends regime types. At the same time, illiberal states are increasingly subverting the norms of global information openness to pursue their own interests, misaligned with the imaginary of the liberal international information order (Farrell & Newman, 2021). 

The association between democracy and human rights is complex, though indexes often conflate these notions. The extent to which they overlap depends on which definition of democracy is used: “Thin or procedural definitions of democracy afford less space for human rights than thicker or social definitions, while it may be possible to conceive of some attributes of human rights sitting outside the conceptual space of democracy” (Landman, 2018, p. 50). Moreover, “studies show that democracy and human rights are indeed positively correlated with one another but not perfectly so” (Landman, 2018, p. 54). The measurement of internet freedom – due to its assumed association with both democracy and human rights – thus defies easy assumptions about appropriate indicators and socio-political relevance. 

In the age of pervasive datafication and state/corporate surveillance (Dencik et al., 2019), both democratic and non-democratic states find themselves dealing with the implications of sophisticated technologies, such as biometric systems and facial recognition, as well as with issues of individual consent and privacy amid the growing extraction of data from the bodies and behaviours of people using communication technologies. Corporate and state actors can also promote a neo-colonial attitude towards networked communication structures and the use of citizen data in commercial or security contexts, downplaying their human rights impact (Couldry & Mejias, 2019; Nothias, 2020).   

In light of the challenges outlined above, it should be noted that the development of internet governance principles and the benchmarking of internet freedom indicators is an iterative process that involves multiple stakeholders, including civil society organisations, private sector initiatives, and (inter)governmental organisations (Hawtin, 2011). In this process the developing definitions and parameters of internet freedom have tended to be dominated by a Western-leaning perspective that has uncritically assumed that democratic values are embedded in the very origins of the internet (Morozov, 2011). Grappling with and rethinking these assumptions requires frameworks that possess analytical capacity when applied to invasive internet policies and practices implemented in democracies, as well as analytical complexity when applied to policies and practices in non-democracies. It is therefore important to trace the evolution of the field of internet freedom rankings and their relationship to the ever-changing notion of internet freedom itself.

Section 4. Methodology

We draw inspiration from the relational approach to international rankings proposed by Beaumont and Towns (2021) and aim to identify the diverse types of rankings and indices measuring internet freedom and their aims and objectives, as well as understand the interrelations among the actors involved in their creation and promotion. Taking the development of rankings related to internet freedom over the period 2002-2020 as our object of study, we approach this rankings landscape as simultaneously the product of and constitutive of globally influential but shifting understandings of internet freedom and reflective of global power relations. While we are interested in exploring specific rankings and their mechanisms for measuring and representing internet freedom, we also attend to the relational evolution of “how [international rankings] are sustained” (Beaumont & Towns, 2021, p. 1475) and how the broader idea of internet freedom is shaped by “the particular competition produced, maintained, and encouraged” by specific ranking organisations (Beaumont & Towns, 2021, p. 1476).

4.1. Selection of rankings

International rankings take various forms. Kelley and Simmons (2019, p. 493) differentiate between indexes or indicators, which “use numbers or grades to rate or rank state performance”, categorical assessments that “use ordinal categories to produce (un)flattering peer groups” and blacklists or watchlists that “draw stark distinctions between compliers and offenders''. Our research identified examples of all three types.

While striving to map the broad landscape of various types of rankings, reports and indexes related to internet freedom (including those currently in operation and those already defunct), we set key criteria for selecting rankings for in-depth analysis. Our initial review identified diverse outputs providing information on internet freedom globally, yet only a selection of these take the form of indexes or rankings that rank, rate, categorise, label or otherwise evaluate actors (Kelley & Simmons, 2019). By doing so, rankings (and by extension the actors producing them) “construct for themselves an authority position in the field of practice, simultaneously clarifying the rules of the game and allocating status” (Beaumont & Towns, 2021, p. 1476). We therefore selected outputs that satisfied the following criteria:

a. Provide regular reports with qualitative and/or quantitative data;
b. Imply certain hierarchies by ranking and comparing actor performance using indicators to assess internet freedom and assigning points or scores;
and/or:
c. Assign categories or labels to groups of actors;
d. And propose value judgements or recommendations based on the numerical rankings, labels or categories assigned.

After compiling the initial database of over 15 projects tracking internet freedom between 2002 and 2020, we selected all rankings that fit the above criteria (a), b) and/or c), and d)) (see Table 1 for an overview of the resulting seven rankings).

Based on the applied criteria, initiatives that publish reports or provide data and/or interactive dashboards, but do not rank or provide narrative or label-based assessment of ranking results, were excluded from our analysis. For instance, while OONI (Open Observatory of Networked Interference, which tracks internet shutdowns and filtering) provides country-level data on internet censorship, it does not rank countries by their track record. Similarly, the Internet Monitor, a project by the Berkman Klein Center for Internet and Society at Harvard University, offers interactive dashboards to review aspects of internet access and freedom in several countries and aggregates data from other research projects but does not provide a ranking or labels for countries assessed. Access Now’s Transparency Reporting Index only provides data on whether or not internet companies release annual transparency reports but does not otherwise rank them or offer label-based evaluations.

4.2. Data collection and analysis

Data was gathered from the official websites of the respective organisations, as these are seen as key communicative devices for a growing majority of rankings (Ringel, 2021a). All editions up to those covering the year 2020 (the year our research was conducted) were included. To collect information about rankings published in earlier years and their methodological complements we often had to resort to using Google Search for editions not explicitly linked from the project main page (searching within the project website to locate PDF files of reports, e.g. using the search query format such as “site:[project website URL] [report title] [report year]”) and the Internet Archive’s Wayback Machine for those editions not available on live websites (again, searching within specific project websites for snapshots from a given year and ranking titles). For instance, Reporters Without Borders’ Enemies of the Internet rankings were not directly linked from the homepage and the integrated search function returned only six of its eleven editions. When available, both the full PDF files of reports and their web presentation were examined.

Because we understand internet freedom rankings as “public performances” (Ringel, 2021a) our assumption is that ranking producers will make the information they view as impactful publicly available and will ensure it is accessible to a range of users on their websites. It was important for us to note the public availability and accessibility of reports, data sets and methodology notes, since Ringel argues that websites increasingly are “at the heart” of rankings’ efforts to reach larger audiences and are “essential” in rankings releases (Ringel, 2021a, p. 69). While, in the course of research or during a journalistic investigation, scholars or reporters might reach out directly to ranking organisations to request additional formation, other groups, such as grassroots digital rights advocates, policymakers, corporate policy advisers or individual online users, are far less likely to go the extra mile, relying instead on the “disclosure devices” immediately available to them online (Hansen & Flyverbom, 2015). During our data collection, we strove to first emulate a typical user experience of online information discovery for each of the rankings, and only if our efforts were not successful, resorted to using internet archival tools mentioned above. We discuss the issue of the public availability of ranking data and archiving practices in more detail in section 5.

For the first step of the analysis, which examined the organisational aspects of the internet freedom rankings landscape, we identified the following aspects for each of the projects (presented in Table 1 in section 5):

1. Organisational backing (producing actor name and type);
2. URL of ranking webpage;
3. Years of publication;
4. Scope (number of countries/companies evaluated and change over time);
5. The rankings’ stated focus;
6. Source(s) of funding;
7. Which of our selection criteria the rankings match.

For the second step of the analysis, which analysed the rankings as “public performances”, we performed a more in-depth coding and evaluation of the rankings’ format and presentation, use of quantitative and qualitative data and other visual and narrative “disclosure devices” (Hansen & Flyverbom, 2015), availability and continuity of data and reports, disclosures about methodology and its limitations. All of these elements contribute to transparency and inclusivity (Ringel 2021b) as strategies that ranking organisations apply to maintain their credibility and proactively counter possible criticism. We also documented other elements of the rankings’ “public performances” (Ringel 2021a), such as advocacy campaigns, media events or calls to action. In analysing all publicly available materials for each of the selected rankings, we coded the following aspects (see Table 2 in section 5):

1. Quantitative data provided on countries/companies;
2. Qualitative data (case studies, incident reports, narrative analysis) provided;
3. Use of information and visualisation tools and devices (charts, maps, dashboards, etc.);
4. Possibility of tracking changes in data over time;
5. Public availability of datasets online, including data for each year of publication (and archival data for rankings now defunct);
6. Public availability of ranking methodology, including changes in methodology for each year of publication (and archival information for rankings now defunct);
7. Use of other “visibility devices”: launch events, press releases, campaigns;
8. Availability of calls to action, recommendations, suggestions.

The next section presents the results of the analysis.

Section 5. Mapping the internet freedom rankings landscape

In this section, we first approach rankings as organisational units (Ringel 2021a) to examine the emergence of internet freedom rankings and their development and persistence over time. We then approach them as “public performances” (Ringel 2021b), drawing attention to their public communicative practices, transparency and stakeholder engagement practices. Synthesising both approaches, we subsequently explain the key changes in the internet freedom rankings landscape over the past two decades by interpreting them against the development of (the understanding of) the internet itself and its societal embeddings. Finally, we provide vignettes of three rankings to illustrate key trends and explain the diversity within the internet freedom rankings landscape: Freedom House’s Freedom on the Net; Reporters without Borders’ Enemies of the Internet; and Ranking Digital Rights’ Corporate Accountability Index.

5.1. Internet freedom rankings as organisations

While in other domains international rankings are also produced by states, International Organisations or for-profit organisations, on internet freedom the ranking organisations (see Table 1) tend to fall within the category of transnational advocacy (Broome & Quirk, 2015, p. 843). Most are produced by civil society organisations and academic institutions, including collaborative efforts between these actors. These include non-profit organisations and advocacy groups (e.g. Freedom House and Access Now), and academic collectives (e.g. the now defunct OpenNet Initiative, a joint project of Citizen Lab at the Munk School of Global Affairs, University of Toronto; the Berkman Center for Internet & Society at Harvard University; and the SecDev Group in Ottawa).

Though they differ in terms of their geographic coverage, frequency, funding sources and thematic scope, these reports or dashboards are similar in that they publish data related to internet freedom (or its components) in countries around the world in a structured narrative, tabular or graphic format. A number of them (e.g. Freedom House’s Freedom on the Net or Access Now’s Keep It On Internet Shutdowns Annual Report) offer additional narrative framing by providing case studies of specific incidents and further opportunities for comparison of national data, such as score tables or country score change graphs. Some newer reports, such as the Ranking Digital Rights’ Corporate Accountability Index, also assess corporate disclosure on internet freedom or track the use of algorithms and automation. 

Table 1: Organisational landscape: rankings of internet freedom matching our criteria

For optimal readability of this table click here.

As outlined above, rankings not only publish data or indicators assessing internet freedom. They suggest how specific indicators should be used to rank or frame the “performance” of states or corporations. These normative aspects of rankings are important as they represent a particular articulation of how internet freedom is understood and operationalised. They also underscore which indicators and framings have more weight in the context of each particular ranking, allowing us to examine their politics. Of the rankings in our sample, few take a comprehensive view of internet freedom, with the exception of Freedom House’s Freedom on the Net (FOTN) which focuses on internet freedom in the broadest sense (including obstacles to access, limits on content, and violations of user rights) and measures it at state level. The now defunct Web Index is another example of a broadly-formulated index that offered a selection of indicators that, alongside internet freedom and openness, also measured internet access and digital inequality. Other rankings focus on specific aspects, such as censorship and surveillance (OpenNet Initiative, which is no longer active, Keep It On, and Enemies of the Internet) or corporate transparency (Corporate Accountability Index). Yet others only assess the impact on specific stakeholders, e.g. journalists (World Press Freedom Index).

5.2. Internet freedom rankings as “public performances”

We identified a striking diversity in the rankings’ approaches to framing internet freedom and their presentation strategies (see Table 2). These, we argue, reflect their diverse aims as well as their intended audiences.

Table 2: Rankings as “public performances”

For optimal readability of this table click here.

5.2.1. Ranking approaches and tools

A classic approach to ranking is that of an “index” (Kelley & Simmons, 2019) – ordering actors based on their composite numerical scores attached to certain indicators of internet freedom. These indicator scores can be high-level (e.g. OpenNet Initiative assigned scores for internet filtering on a five-point scale in four broad categories: political, social, conflict/security and internet tools) or more granular (e.g. Freedom on the Net groups questions into three categories with a maximum possible point-based score in each). Six of the seven rankings include some form of numerical indexing (with the exception of Enemies of the Internet).

Rankings can also assign labels or categories, creating a hierarchy of actors who perform better or worse in specific areas. These labels can be included in each of the indicators or reflect the sum of all numerical points achieved by a particular actor. Four of the seven rankings make use of labels or categories. For instance, OpenNet Initiative provides five labels to measure the level of internet filtering (from “pervasive” to “suspected filtering” or “no evidence of filtering”), while Freedom on the Net labels countries as “free”, “partly free” or “not free”. While most rankings combine numerical and category-based ranking, RBW’s Enemies of the Internet is the only ranking relying on a purely qualitative labelling approach that emphasises framing actors rather than ranking them based on a particular weighing of indicators. 

The rankings use various tools that facilitate deeper analysis, such as dashboards enabling comparison between actors, across years or on specific indicators; static or interactive maps inscribing the ranking outcomes into geopolitical borders; narrative elements such as in-depth case studies, testimonies or quotes; and supplemental materials targeting particular audiences, such as data sets or additional graphics. Graphic presentations are increasingly popular as they allow for an easier-to-grasp and more impactful presentation of ranking results. For country rankings, map-based visualisations remain the most popular and are used by many of the rankings (see Table 2). Bar charts and colour-coded visualisations assigning certain colours to labels or indicator values are also used extensively.

Publicly available methodology guides also help shed light on data collection and ranking practices. The explicit disclosure of the approach and methods used in the ranking (which six of the seven rankings provide, as seen in Table 2) can serve as a corrective when assessing the relative importance of subjective expert judgments vs “objective” indicators; data aggregation methods; ranking validity (does indicator/data capture what it purportedly measures); the possible overlaps between indicator components; the replicability of the ranking’s methods; and the overall interpretation of what constitutes internet freedom.

5.2.2. Presentation and structure of ranking reports

The ranking reports are structured in several typical ways. They begin with a narrative summarising the key findings of the report, illustrated by specific data or examples. While the Web Index, for instance, only provides a token narrative section, others present the narrative as a key element of the report that complements the other structural parts by connecting key indicators, explaining change in any given period or providing evidence in the form of examples or case studies. Freedom on the Net, for instance, provides an overview text as well as country-specific narrative reports. Enemies of the Internet relies predominantly on an overarching narrative and case studies to single out key offenders violating digital rights globally.

Other structural elements of rankings we encountered include lists of criteria, labels imparting certain qualities or numerical values attached to key indicators. These are typically seen as more objective elements that allow for direct comparison between actors, but they usually require careful documentation and methodological explanations to support their validity. As such, they are almost never used on their own and usually are accompanied by corresponding narratives or visualisations. As mentioned above, graphics and dashboards are commonplace devices for making numerical data more accessible. At their conclusion, ranking reports typically feature the sources used in their research and analysis as well as provide links to raw datasets or supplementary appendices.  

The ranking organisations engage in extensive adjacent actions such as launch events, press releases and media campaigns to support ranking visibility and to engage intended audiences. These range from communications targeting specific stakeholders (e.g. media, policymakers or investors) to hybrid report launches accessible to a broad range of interested parties.

5.3. Explaining key trends in the development of the internet freedom ranking landscape

The diversity of internet freedom indexes demonstrates the complex evolution of the field characterised by several key trends. The first trend is greater granularity: while some of the older rankings, such as FOTN (2009) or Web Index (2012), assessed levels of internet freedom in a comprehensive sense, many of the rankings introduced later, e.g. RDR’s Corporate Accountability Index (2015) or Access Now’s Keep it On (2016), focus on specific aspects of internet freedom, such as transparency or internet shutdowns. Reporters Without Borders’ approach has undergone a similar transformation: the World Press Freedom Index (born in 2002) has a broad focus on media freedom, including internet freedom, whereas Enemies of the Internet (launched in 2005) focuses mostly on online censorship and surveillance. 

The shift to greater granularity is also reflected in the unit of assessment: as the indexes evolve there is a gradual move from broad country-level assessments to scrutinising corporations or particular state agencies, institutions, groups or individuals. Especially in ranking narrative reports there is a more nuanced depiction of actors impacting internet freedom (e.g. in Russia it’s not just “the Russian authorities” but Roskomnadzor, the agency overseeing the internet and media sector). This exemplifies a more sophisticated understanding of which actors influence internet freedom in national or global contexts. The same is true with regard to the granularity of assessing who is impacted by particular regulations or restrictions on internet freedom. For instance, RWB’s early Enemies of the Internet reports speak of “netizens” who are “fighting back” or “imprisoned” for their efforts without detailing who they are. However, in 2013 the report differentiated between “netizens jailed” and “journalists jailed”, using more precise categories of actors.

The second trend concerns the rankings’ increasing complexity, both in terms of methodological approaches and in terms of presenting their results. In part, this is because the internet itself has evolved, thus requiring new vocabularies and categories of assessment. The complexity also reflects the growing sophistication of state and non-state actors seeking to define, regulate or limit internet freedom. On the one hand, as was argued above, internet freedom now has many more constituent parts and stakeholders. On the other hand, the tools for “information control” (Deibert et al., 2010) have also become more sophisticated, requiring a constantly evolving set of definitions for what qualifies as an attack on internet freedom. In response to this, the internet freedom indexes have also drawn on greater experience and professionalisation in the field at large and have expanded their methodology by adding new categories of assessment (e.g. Corporate Accountability Index has begun to assess transparency of corporate algorithmic practices). Others have changed in a less linear manner: RWB has diversified its Enemies of the Internet report to include not only states, but also corporations and individuals. The presentation of the results and their implications has also changed, with multiple rankings using more complex dashboards, interactive maps and other tools for tracking change over time and displaying country/company results in a global context.

This increasing complexity, however, has not consistently resulted in greater transparency and accessibility of ranking methodology or data integrity. Though some rankings have consistently provided access to their methodology and underlying data from the start (e.g. Access Now’s Keep It On Internet Shutdowns Report), others relegate methodological notes to appendices, making them less visible. This matters as rankings are “likely to exert pressure” on the status quo only if they are “released to large audiences - publics” (Kelley & Simmons, 2019 in Ringel, 2021a, p. 55). While dashboards and infographics or case summaries in the main body of the annual reports typically draw more public attention as “visibility devices” (Hansen & Flyverbom, 2015), appendices are easily overlooked. Both active and defunct rankings also perform poorly on (publicly) archiving past ranking data and the reports themselves. The landing page for the 2008 Enemies of the Internet ranking is no longer available online and exists only as an Internet Archive snapshot, while Web Index’s data and graphics are only partially available online, though the project has made an effort to keep the archive website alive. This lack of a systematic approach to data retention or tracking methodological changes may constrain the capacity of ranking organisations to wield their influence as “referees” (Beaumont & Towns, 2021) when making policy-related or rights-related claims that require reliable (and publicly available) supporting evidence.

The third trend is a shift in inclusion criteria and framing, illustrative of the gradual change in perception of the concept of internet freedom and the growing field of actors impacting its dimensions. In indexes of both media freedom (RWB’s World Press Freedom Index) and internet freedom (Freedom House’s FOTN), there is not only a change over time in the number of countries included, but also a shift from focusing predominantly on autocracies to also including democracies. This signals a gradual change in how ranking organisations – and popular opinion – perceive who poses a threat to internet freedom. Whereas early (Western) ideas of internet freedom were mostly predicated on the dichotomy between autocracies (threatening internet freedom) and democracies (defending it), this framing has gradually expanded to acknowledge a more complex reality in which the actions of democratic governments should similarly be scrutinised. More recent rankings have also begun acknowledging the actorness of corporations with regard to internet freedom and initially scrutinised Western technology companies in recognition of the outsized power they wield in shaping internet freedom globally. 

In the next section, we build on this overview to provide in-depth vignettes of three internet freedom rankings to illustrate the key trends and highlight the diversity among internet freedom rankings.  

5.4. Ranking vignettes

5.4.1. Freedom House’s Freedom on the Net 

Modelled after the “classic” democracy and media freedom indexes, Freedom on the Net (FOTN) ranks country performance on the basis of a set of indicators (Kelley & Simmons, 2019), presenting itself as an authoritative record of the current situation based on expert research. The ranking was first released in 2009 to complement Freedom House’s other rankings: Freedom in the World (published since 1973) and Freedom of the Press (1980-2017) and was pitched as “a ranked, country-by-country assessment of online freedom” (Freedom House, 2021a). It includes ranked numerical scores, categorical labels and in-depth country reports, as well as a global overview of key trends. Since 2009 the number of countries in the report has increased from 15 to 65. 

Freedom House is a non-profit non-governmental organisation founded in 1941 in the United States. It describes its mission as “founded on the core conviction that freedom flourishes in democratic nations where governments are accountable to their people” (Freedom House, 2021b). The organisation’s rankings and reports “frame the policy debate in the United States and abroad on the progress and decline of freedom” (Freedom House, 2021b). Though the organisation operates independently, it receives substantial support from the US government: in 2006, 66 percent of its funding came from US government grants, while in 2016, this figure rose to 86 percent.

The FOTN report’s main aim is to produce “research and analysis”-based regional assessments of internet freedom, while also facilitating “fact-based advocacy” and “capacity-building” through the release of side reports and public events (Freedom House, 2021b). The methodology, a survey designed to measure the state of internet and digital media freedom in each country, comprises 21 questions and almost 100 subquestions, falling into three categories: 

1. obstacles to access: infrastructural and economic barriers to access, state efforts to block specific applications or platforms, legal controls and ownership of internet and mobile phone providers;
2. limits on content: filtering and blocking of websites and other forms of censorship and self-censorship, manipulation of content, the online news media landscape, and the uses of digital media for civic and political activism;
3. violations of user rights: legal protections and restrictions on online activity, surveillance and privacy limitations, and sanctions for online activity, such as legal prosecution, imprisonment, physical attacks, or harassment. 

Each question is assigned a maximum numerical score, and the scores from the three areas are combined into a total country score between 0 and 100. Countries are then labelled "free" (100 to 70), "partly free" (69 to 40), or "not free" (39 to 0) based on the total scores, placing countries into “peer groups” (Kelley & Simmons, 2019). Charts representing various aspects of the scoring and a map colour-coded by country status has been included in the reports from the start. Since 2016 an interactive version has been available online to account for the increasing complexity of internet freedom measurements.

The FOTN team works with researchers to produce country reports and scoring, training them in the methodology and convening regional review meetings to discuss scores. Rankings are reviewed on an individual and a comparative basis and compared with the previous year's findings. Narrative reports presenting structured responses to key survey questions complement numerical scores and provide evidence or case studies to explain why the scores have (not) changed.

During its relatively long existence FOTN’s ranking approach has invariably changed. The number of countries assessed has gone from 15 to 65, so a full longitudinal comparison is only possible for a smaller subset of states. Though the number of key questions has remained relatively stable, some have been added or rephrased. For instance, in 2014, question 7 in the section Violation of User Rights read “Are bloggers, other ICT users, websites, or their property subject to extralegal intimidation or physical violence by state authorities or any other actor?” In the 2020 version, it reads: “Are individuals subject to extralegal intimidation or physical violence by state authorities or any other actor in retribution for their online activities?”.  

Earlier versions of the FOTN methodology cited a clear conceptual basis for its approach to internet freedom. The 2014 methodology document states:

Freedom House does not maintain a culture-bound view of freedom. The project methodology is grounded in basic standards of free expression, derived in large measure from Article 19 of the Universal Declaration of Human Rights [...].This standard applies to all countries and territories, irrespective of geographical location, ethnic or religious composition, or level of economic development.

The document also recognises that “in some instances freedom of expression and access to information may be legitimately restricted”. In contrast, the 2020 version of the methodology does not outline FOTN’s politics with regard to defining internet freedom.

Over time, FOTN has included more Western democracies in addition to the “usual suspects,” including autocracies and former Soviet states. The 2014 and 2020 methodology documents also address the challenge of country-based scoring, recognising that governments are not the only actors impacting internet freedom in a given geographic location. The 2014 document notes that “pressures and attacks by non-state actors, including the criminal underworld, are also considered”, while the 2020 methodology page instead includes “actions by nonstate actors, including technology companies”.

FOTN demonstrates how ranking organisations have to negotiate necessary methodological adjustments with maintaining credibility based on longitudinal consistency (Ringel, 2021b). While methodological changes are acknowledged and specified in the appendices, they are not highlighted in the public presentation of the rankings. Only select challenges have been noted in the narrative reports; for example, since 2014, Crimea and the parts of eastern Ukraine occupied by Russia-led forces are not assessed in the main Ukraine report, as these territories are not under government control. 

With regard to public archiving of annual reports, raw data and methodology documents, we identified a similar balancing between the degree of transparency expected by the organisation’s stakeholders to maintain credibility and operational capacity. At the time of the study the FOTN methodology was only available for 2014-2020. While all past reports were available from the FOTN main page, the 2009-2015 reports were only available as PDF files containing the overview essay and country scores and, in a few cases, the country reports. Country-specific webpages only linked to reports after 2016. Raw data and score sheets were only publicly available for the two latest editions.

As an established non-profit organisation known for its international rankings in the related domains of democracy and press freedom, Freedom House was able to draw upon its reputation and organisational capacity to establish an authoritative ranking on internet freedom. It uses a well-crafted suite of additional public outreach tools to frame and promote its internet freedom ranking, including public launch events, press materials and advocacy campaigns. While FOTN provides considerable transparency about its methodology and assessment approach, and presents its findings in an accessible way, the limited public visibility of archival materials could undermine the usability of the ranking for audiences that expect access to comprehensive ranking materials as evidence of longitudinal trends (Ringel, 2021a, 2021b).

5.4.2. Reporters Without Borders’ Enemies of the Internet

“Blacklists” or “watchlists” similarly engage in selecting and categorising actors but prioritise the framing of the selected list over executing a comprehensive and systematic comparison (Kelley & Simmons, 2019). Enemies of the Internet highlights actors involved in excessive restrictions of internet freedom or digital surveillance. Its core aims include raising awareness and focusing the attention of policymakers and civil society on specific issues and countries. This allows for a greater gap between what is “measured” and what is presented: the ranking has a clear central message to convey, while the presentation of factual data takes a back seat. Recognisant of its intended audiences and the role of public opinion in promoting political action it prioritises concise narrative reports and press releases along with visualisations that quickly convey their key message.

Published annually between 2005-2014, the Enemies of the Internet index reemerged in 2020, now renamed Digital Predators. The first edition was prepared as a report for the World Summit on the Information Society in November 2005, after which it continued as a public-facing campaign. The publication date shifted from November to 12 March in 2008 as RWB launched the International Online Free Expression Day, later renamed the World Day Against Cyber-Censorship. The publication date is the index’s most consistent element. Throughout the years, it has shifted from listing countries (2005-2012; between ten and fifteen “enemies”) to countries and corporations (2013; five each) and finally, corporations and state institutions (2014; 2020; thirty-one and twenty, respectively). Signalling a shift in inclusion criteria, the 2008-2012 editions also identified additional “countries under surveillance”, while the 2005 edition included an additional ten “countries to watch”. 

By selecting countries, corporations and institutions and branding them as “enemies”, the ranking, by implication, compares them (negatively) to other entities that are not listed. While RWB does not explicitly rank selected actors, their presence on the list signals they are among the worst offenders. The reports often use combative language and pit “netizens” in direct opposition to (authoritarian) governments. The 2010 report, for example, claims that “[t]he outcome of the cyber-war between netizens and repressive authorities will [...] depend upon the effectiveness of the weapons of each camp” (RWB, 2010). Watchlists and pejorative labelling (“enemies”; “predators”) (Kelley & Simmons, 2019) are used by the ranking to frame the actions and policies of particular countries or companies as harmful and in need of intervention. 

This framing, however, is potentially problematic since the methodology for the selection is not disclosed. The closest the reports get to clarifying the selection criteria are brief characterisations of the set of “enemies”, which change significantly from one year to the next. For example, in 2009, the index says they “have all transformed the network into an intranet, preventing Internet users from obtaining news seen as ‘undesirable’” (RWB, 2009, p. 2), while the 2010 report refers to them as the “worst violators of freedom of expression on the Net” (RWB, 2010). The 2012 report calls out countries that “combine often drastic content filtering with access restrictions, tracking of cyber-dissidents and online propaganda” (RWB, 2012, p. 10). The reports combine a narrative describing key annual trends (often referencing countries not included in the list, including democracies) with country files summarising key events for each “enemy”. How these country files were compiled or on what basis is not explained. Country files also vary in structure and this lack of standardisation further hampers comparability (both between actors and for each actor over time). 

The lack of a transparent methodology and consistent presentation means the reports effectively are a collection of illustrative anecdotes, aimed at supporting the objectives of the associated campaign. The fact that several reports have a theme reinforces this. The 2013 report was presented as a “special edition” focused on surveillance while the 2014 aimed to highlight the role of particular entities, because “identifying government units or agencies rather than entire governments as Enemies of the Internet allows us to draw attention to the schizophrenic attitude towards online freedoms that prevails in some countries” (RWB, 2014, p. 3). The latter was meant to refer, in particular, to the American NSA and British GCHQ post-Snowden and the export of surveillance technologies to autocracies. 

Similarly to FOTN, Enemies of the Internet’s role in the internet freedom ranking landscape draws on the organisational capital established by RWB’s other long-standing – and more methodologically robust – index, the World Press Freedom Index. While RWB draws on its reputation as a defender of free expression to attract public attention to ranking results and promote advocacy messages to stakeholders, the Enemies of the Internet ranking lacks the methodological rigour and transparency of Freedom on the Net. Its authority, if recognised, relies fully on that of the already established press freedom ranking that has such methodological grounding. Despite methodological weakness, Enemies of the Internet uses the commonly identified ranking strategies of “blacklisting” and assigning “peer group” labels to actors in the digital ecosystem and relates their narratives to key trends in the internet freedom landscape to achieve their advocacy goals.

5.4.3. Ranking Digital Rights’ Corporate Accountability Index

In contrast to the preceding two rankings, the Corporate Accountability Index ranks corporations rather than states. It explicitly foregrounds impact without making concessions on the methodological rigour and transparency of its ranking. A key difference is the direct involvement of the ranking organisation in suggesting specific improvements to those being ranked rather than relying on other advocacy organisations (as with the previous two rankings). Its core aims lie in connecting directly with those they rank, using evidence-based indicators to persuade these actors to change their practices.

The Corporate Accountability Index is produced by Ranking Digital Rights, an independent project housed at New America Foundation and affiliated with the Open Technology Institute. RDR receives funding from a variety of non-profit and charity foundations as well as the US Department of State. The ranking assesses telecommunications and digital companies’ transparency and accountability and aims “to promote freedom of expression and privacy on the internet by creating global standards and incentives for companies to respect and protect users’ rights” (Ranking Digital Rights, 2021). 

The ranking’s pilot edition was launched in 2015 and evaluated 16 prominent internet and telecommunications companies, including Google, Facebook and Microsoft, on their public commitments and disclosed policies affecting users’ freedom of expression and privacy. By 2020 the ranking had expanded its geographical reach to include 26 companies from Russia and China to the US and South Korea. 

The methodology includes 58 indicators grouped into three key categories: governance, freedom of expression and privacy. The categories are grounded in internationally recognised agreements, such as the ​​U.N. Guiding Principles on Business and Human Rights, and terms used in the indicators, e.g. “user information” or “disclosure,” are defined explicitly. Changes in the methodology over time are documented and explained and the methodological guidelines are publicly available. Engagement with the corporations ranked is part of the index’s design: soliciting feedback from companies is incorporated as one of the seven research steps.

The index’s indicators assess corporate practices and disclosures and assign scores based on whether there is evidence of full or partial disclosure on specific issues, from explicit commitments to free expression to explanation of how user data is stored and shared. The ranking assesses each company as a whole, as well as several of its most popular services or applications. As with Freedom on the Net, over time the ranking’s indicators have evolved to include new concerns. The number of indicators went from 35 in 2019 to 58 in 2020, expanding categories and adding new indicators assessing company policies on targeted advertising and algorithms. The scoring is performed by independent researchers trained in the RDR methodology, reviewed by a second set of peer researchers and cross-checked by the RDR team, who also liaise with company representatives to solicit feedback.

The Corporate Accountability Index allows for comparisons of indicator scores year-on-year but acknowledges the limitations in light of methodology changes. For each report year the ranking provides a detailed methodology document outlining key indicators, scores and the scoring process. E.g. the methodology for the 2020 report explains the expansion of the indicator lineup and the resulting changes in the score logic, noting that “changes to the methodology resulted in significant score declines for most companies in our ranking”. 

In presenting its annual ranking the Corporate Accountability Index uses interactive dashboards and graphics. Narrative reports are used to summarise the overall results and highlight key issues, whereas company-specific “report cards” combine infographics with a summary of key takeaways. Raw data are publicly available for every annual report, along with other materials such as key findings. The 2017, 2018 and 2019 reports also provide a map with all the ranked companies. 

The ranking reports are complemented by a suite of targeted communications, including public launch events, digital rights advocacy events in collaboration with other groups, thematic policy briefs and company scorecards aimed at the corporations ranked. The ranking makes detailed recommendations for improvement and points out priority areas. It also makes recommendations to investors using responsible investment practices as a lever to exert pressure on corporate actors. The ranking organisation’s efforts to directly engage with the corporations it ranks appears to go beyond its role as “referee” (Beaumont & Towns, 2021) – assessing corporations’ performance – and resembles the role of “coach” – suggesting to corporations how to perform better in the ranking. Yet, contrary to how ranking “coaches” typically operate, Ranking Digital Rights does not act out of commercial interest (further research can help clarify their activities in this domain).

The Corporate Accountability Index defines internet freedom more narrowly in accordance with internationally accepted human rights norms, exemplifying the trend for greater granularity. Displaying methodological complexity and transparency, its emphasis on the activities of large corporations also demonstrates a shift in inclusion criteria. The approach to evaluation using granular indicators and scoring is resource-intensive but can potentially have more impact on the actors it seeks to influence through proactive outreach and stakeholder-specific recommendations.

Section 6. Conclusion

In this article, we have traced the development of internet freedom rankings over the course of two decades. Our aim was to understand how and why such a complex landscape of rankings has emerged and to identify how the varying approaches to rankings have played a role in defining internet freedom and presenting it as a salient issue. We found that, in many respects. producers of internet freedom rankings perform similarly to international ranking organisations in other fields, employing the strategies of “indexing”, “labelling” or “blacklisting” to evaluate actor performance and adjusting their “public performance” to establish authority with intended audiences. We have also established that, over time, the field of internet freedom rankings has developed towards greater granularity and increasing complexity and displays shifts in inclusion criteria, accounting for a new reality where the actions of both autocracies and democracies (as well as other non-state actors) can potentially threaten internet freedom and require greater scrutiny. The number and diversity of rankings, as well as the extent of their continuous adjustment and the persistence of some rankings despite the demise of others, show that internet freedom remains a contested concept and a “moving target”: the internet itself and its interwovenness with social and political life is changing constantly, and the boundaries to online freedoms and rights are continually renegotiated.

As our research shows, understanding these shifts requires us to look beyond the rankings as a set of “disclosure devices'' to the organisational capacity and history of the ranking organisations themselves, which equally impact their authority as arbiters or “referees” of internet freedom. For example, our findings on the publishing of complementary materials (datasets, methodological appendices) and the (public) archiving of the rankings confirm that, as Ringel (2021b) suggests, the activities of ranking organisations in the area of transparency are driven by their need to maintain credibility (as in the case of FOTN and Corporate Accountability Index). However, when introducing new rankings, ranking organisations can also rely on their existing reputation and, as a form of preemptive defence against external criticism, tailor their performance on this front to what their intended audiences (and likely critics) expect of them. This tendency does not necessarily lead to more methodological precision (as illustrated by Enemies of the Internet).

The inconsistent public archival practices of the internet freedom rankings we examined, and the varying degrees of transparency about methodological changes, warrant a critical examination of their utility as sources of longitudinal trends in the field of internet freedom and as drivers of public debates about the nature and severity of threats that state or corporate power may pose to a free and equitable digital ecosystem. As ranking organisations navigate capturing the moving target of internet freedom they have necessarily made methodological adjustments that affect the rankings’ consistency over time. Yet, as narrativisation takes precedence in public communication, it is precisely these changes over time which are emphasised (rising or falling in the ranking). This warrants a critical reevaluation of how, and for what purposes, these rankings can be used by scholars, policymakers and the public-at-large.

Collectively, the rankings have successfully helped put internet freedom on the map as an issue of global import. However, how the respective ranking organisations relate to each other and shape approaches to internet freedom assessments has implications for the developing set of issue definitions and proposed solutions.

Our analysis of the rankings landscape over time shows that non-profit ranking organisations have more persistent presence, while rankings developed by academic collaborations (e.g. OpenNet Initiative) have a shorter lifetime, most likely due to structural constraints and limited funding support (since academic funding tends to be project-based). This pattern has implications for the ongoing development of the ranking methodologies, as well as for the diversity of the field of actors contributing to the definition of internet freedom, and requires further research.

Finally, our analysis demonstrates how dominant frameworks for assessing internet freedom continue to be predicated on democracy- and media freedom-derived indicators, in part due to the organisational histories and the neoliberal background of the leading ranking organisations that produce them. Newer players in the field of internet freedom rankings are beginning to push back against the dichotomy of ‘democracy/non-democracy’ and instead work towards a more inclusive approach, acknowledging that the complex issues associated with internet access, internet infrastructure, and internet governance cut across regime types, involve multiple stakeholders and extend beyond the media sphere. Nonetheless, these more recent rankings similarly approach internet freedom from a Western perspective centred around the notion of individual liberty. Consequently, civic and political rights tend to receive the greatest emphasis across all rankings, while the relation between the internet and the realisation or violation of socio-economic rights features less (as demonstrated by the demise of the Web Index). This may have implications for both advocacy focus and policymaking priorities driven by the rankings’ findings and narratives. Further research should examine the (possibly mutually reinforcing) correlation between the evolving definition of internet freedom shaped by the rankings and the shifts in the internet governance landscape.