Ofcom has recently carried out a consultation into the future of media plurality in the UK that includes a review of current media ownership rules. It’s clear that the current regulatory framework is insufficient to meet the challenges of an ever more concentrated news market, especially in local print news, online news and more generally at the level of wholesale newsgathering across platforms.

The system needs not gentle tinkering but some major modifications if it is to adequately cater to diverse audiences and to support media content in the public interest.

Total brand reach (Sept 2020) Source: PAMCo

I welcome Ofcom’s proposal to extend the scope of the public interest test framework and to retain the remaining rules regarding cross-media ownership and restrictions on media ownership in respect of political bodies and local authorities. However, there are two further and urgent changes needed to the existing plurality framework.

First, it is essential that Ofcom is empowered to carry out regular plurality reviews. These should take place every four years, in order to account for organic and dynamic market changes that can impact on the extent of plurality, and to address significant concentrations over recent years that have been well documented.

Second, these reviews should include the monitoring and assessment of major algorithms in respect of plurality goals. The current draft Online Safety Bill presents a timely opportunity to elaborate and define plurality goals as they should apply to major news algorithms, given that Ofcom is already tasked with proposing a code of practice for tech companies in respect of the bill’s broad stipulations. This is the most practical, proportionate and logical approach to factoring in online intermediaries into plurality regulation.

The need for regular reviews

There is compelling data and evidence to suggest that the existing news media landscape falls significantly short of Ofcom’s plurality objectives. In particular, the public interest test framework fails to capture concentrations of power that can and have developed ‘organically’, i.e. as a result of dynamic changes in technology and markets rather than as a result of merger activity. For instance, since abandoning its paywall in 2015, the online edition of the Sun newspaper saw its reach grow at an unparalleled and unprecedented rate, becoming the largest online news provider in the UK within three years. This reflected a significant consolidation both within the online news market and across platforms which has not triggered a regulatory review due to falling outside of the current framework.

Combined weekly circulation by publisher (2020)

Research from the Media Reform Coalition shows that in 2015, three companies controlled 71% of national newspaper readership; by the end of 2018, the same three companies – Rupert Murdoch’s News UK, DMG Media (publisher of the Mail titles) and Reach (publisher of the Mirror titles) accounted for 83%; by the end of 2020 this dominance had increased to 90% of the national newspaper readership. By themselves, News UK and DMG dominate over 70% of the market share of national newspapers. In terms of overall audience reach, drops in circulation of their print titles has been eclipsed by the growth of their online presence: In 2021, the Sun and Daily Mail alone accounted for nearly 38% of total daily UK newsbrand reach. As with print circulation, just three publishing companies account for four-fifths of total brand reach amongst the major print titles: DMG Media, including the Evening Standard (34%), News UK (26%), and Reach Plc (21%).

It is equally clear from the available evidence that the reach and dominance of print newspapers online has not been eclipsed by new entrants. So, while left-leaning non-profit website The Canary attracted 645,260 UK visits in January 2021, this was less than 0.5 per cent of the traffic to the Guardian online, which saw 146.5 million visits. During the 2017 General Election campaign period, the Canary outperformed the Daily Mail online when it came to Facebook shares of articles on the two main party leaders. Yet its actual reach was, over the same period, a tiny fraction of the Mail’s (1 per cent versus 36 per cent). And in the month the election was called, page views of the Daily Mail website outnumbered those of the Canary by a factor of more than 700. Similarly, in January 2021 the right-wing outlet Guido Fawkes saw 2,255,198 visits to its site, approximately 1.5 per cent of that of the Daily Mail with 149 million visits.

Local newspaper titles by publisher (February 2021)

We are seeing similar patterns of concentration in the local newspaper market where just six conglomerates now account for 84 per cent of all titles, while 50 smaller publishers make up the remaining 16 per cent. Back in 2015, six companies dominated the market but with the purchase of Local World by Trinity Mirror (now Reach), the market is even more compressed.

Radio audiences remain healthy, but this is far from a diversified sector. The BBC has a 49.7 per cent audience share while commercial stations have 47.8 per cent, leaving a very small slice of 2.5% for community radio. Once again, however, the commercial sector is highly concentrated with just two large companies, Global and Bauer, accounting for almost 70 per cent of the entire local commercial analogue market and 60.5 per cent of national commercial digital radio. Following the launch of Times Radio in June 2020, Wireless Group (a subsidiary of News UK) now operate four UK-wide talk radio and news stations – more than any other radio broadcaster, including the BBC. New DAB licences have not facilitated greater diversity and local accountability. Rather, we are seeing the consequences of mounting concentration: in February 2019 Global took a decision to scrap around 40 local breakfast shows and to replace them with just three nationwide programmes, resulting in significant numbers of redundancies.

As for television, which remains the most used and trusted platform for news in the UK, this continues to be dominated by just three providers at the wholesale level: BBC, ITN and Sky. In 2020, these three providers accounted for 99 per cent of national/international news (of which the BBC accounted for 65 per cent).

These data underscore the need for plurality reviews outside of merger activity to not only protect against further reductions in plurality, but importantly to address existing deficits as outlined above. It is difficult to imagine that these levels of market concentration are consistent with Ofcom’s plurality objectives, or indeed the statutory framework. As the CMA’s 2018 report on the Fox/Sky merger makes clear, these objectives go well beyond competition concerns.

The need for algorithm governance

Much of the scholarly literature has rightly rejected claims by intermediaries to be entirely neutral conduits through which users find, select and consume news and information. However, this has led in some cases to a presumption that they play an editorial role equivalent to, and directly comparable with, that of conventional news sources. Since intermediaries by definition do not produce original news content, or carry the news produced by an exclusive wholesale provider, the key question from a plurality perspective turns on which wholesale or retail news sources achieve more or less salience on these platforms.

Ofcom’s present approach to assessing plurality on the one hand acknowledges the distinct impact on plurality that intermediaries may have in this sense, by classifying them as news ‘gateways’ as apart from wholesale or retail news sources. But when it comes to analysis of news consumption, intermediaries still tend to be treated as sources alongside wholesale/retail providers. This presents a risk of skewing the data in either of two directions. If, for instance, a user gets their online news predominantly via the BBC website and the BBC News Facebook page, and cites the BBC and Facebook as sources in a survey response, this will be reflected in Ofcom’s data as two separate news sources, effectively double counting the BBC. On the other hand, if they get their news from the BBC website plus various Facebook pages belonging to Sky News, Daily Mail, Guardian, etc then they might still cite BBC and Facebook as their two main sources for news, with the effect that Ofcom’s data would significantly understate their actual news diet.

This resonates with anecdotal evidence suggesting that intermediaries are increasingly prioritising established news sources in both their algorithms and manual interventions into the flow of news across their networks. In 2013, Google’s patent application for its updated news algorithm made explicit reference to the perceived quality and importance of large scale, mainstream news sources. In 2016, it was revealed that Facebook had hired a team of journalists with instructions to ensure that headline stories on mainstream news media were given a ‘boost’ if they weren’t trending on Facebook ‘organically’. As for Apple News, a 2018 study of sources recommended by its editors found that these were dominated by ‘a few major newsrooms’. With Facebook introducing its news service in the UK in January, which involves substantial payments to leading publishers, the prioritising of mainstream content is likely to increase yet further.

This suggests that the most appropriate way to factor intermediaries into both the assessment and regulation of plurality is to focus on the degree to which their algorithms promote a plurality of news sources at both the wholesale and retail level. There are now a number of ways in which it is possible for Ofcom to collect meaningful data on this including but not limited to:

  • Soliciting relevant data from intermediaries themselves
  • Commissioning bespoke data from commercial media measurement and analytics services
  • Carrying out ‘reverse engineering’ and other experimental research designs aimed at determining the range of news sources promoted by major algorithms for various user profiles and/or search queries
  • Adding detailed sub-questionnaires to online news consumption surveys

Importantly, this research should be aimed at assessing plurality both at the aggregate level (i.e. the range of news sources promoted overall) and at the exposure level (the range of news sources to which users are, on average, exposed).

A call for meaningful plurality reviews

We need regular and meaningful plurality reviews that include behavioural or structural remedies as appropriate, together with the regulatory will and power to issue digital intermediaries with specific orders in respect of their algorithm metrics, should their performance fall short of standards as elaborated in the relevant code of practice. Ofcom should also be granted enforcement powers in respect of such orders, allowing the imposition of substantial fines in the event of non-compliance.

Together, these proposed changes will address existing concentrations in wholesale and retail news markets, protect against further reductions in plurality resulting from organic market or technological changes, and, above all, bring online intermediaries into the framework in a way that reflects their particular role and potential impact on news plurality. Since Ofcom first recommended regular plurality reviews to the government in 2010, the idea has attracted consensus support among policymakers. As the gaps in the existing framework become ever more apparent, it is vital that Ofcom revives the proposal whilst factoring in the ever more important contribution of digital intermediaries.

This post is based on research that is part of the UK component of the Center for Media, Data and Society’s Media Influence Matrix, set up to investigate the influence of shifts in policy, funding, and technology on contemporary journalism. The UK component is coordinated in partnership with the Media Reform Coalition and Goldsmiths, University of London and is funded by the Joseph Rowntree Charitable Trust. It is due to report in Autumn 2021.

This post originally appeared on the LSE Media Project Blog and is reproduced with permission and thanks