Algorithmic supremacy? How personalised news is shaping politics

A recent study led by Dr Lin Hu from the Australian National University (ANU) College of Business and Economics (CBE) has explored the effects of personalised news on public opinion and the trending polarisation of politics.

In a world awash with quintillions of data bytes, humans can no longer keep up with the overwhelming volume of information at their fingertips.

As a result, consumers are increasingly turning to personalised news aggregators, algorithm-driven services that analyse individual preferences to recommend content. These platforms, run by infomediaries like Google News, vie for human attention to maximise their revenue.

In light of this growing trend, Lin is exploring how the personalisation of news may be affecting political decision-making, with a particular focus on policy polarisation – a phenomenon that has intensified over the past 40 years.

“Policy polarisation is marked by political candidates taking extreme positions to secure electoral gains,” says Lin.

“This political attitude is notably stark in the U.S. and can fuel societal division, lead to legislative stalemate, hamper democratic processes and degrade mutual respect.”

While previous experts have hinted at news personalisation’s possible effects on polarisation, Lin’s team is the first to examine its influence on rationally inattentive voters (NARI) – individuals who have limited attention and consume news to improve their voting decisions.

Using game theory to simulate electoral dynamics, Lin’s research has revealed that, when exposed to personalised news, NARI voters can actually support the continued existence of polarisation.

“We have discovered that any voter group, from moderate to extreme, can be essential in sustaining a greater degree of policy polarisation,” she explains.

Our findings are groundbreaking because they challenge the traditional view that candidates naturally gravitate towards centrist voters, highlighting the significant impact of infomediaries in shaping political landscapes.

“Unlike traditional infomediaries such as radio and tv, personalised infomediaries can tailor messages to meet individual viewpoints. This is giving voter groups across the political spectrum more power than ever, as any of them could be critical in upholding a candidate’s ideological agenda.”

By discussing the vital role modern infomediaries play in shaping consumer’s beliefs, Lin and her team are also feeding into the debate surrounding the regulation of tech giants, analysing whether limiting access to personal data can reverse the negative effects of personalised news.

In Australia, the news aggregation debate has gained traction after Meta recently announced it would no longer continue to pay Australian publishers for featuring their content.

This development follows the introduction of the News Media Bargaining Code in 2021, which mandates big platforms like Google and Facebook to remunerate media businesses for displaying their news.

“We examined a monopolistic infomediary to capture the market power wielded by tech giants and found the social benefits of many regulatory proposals are not clear-cut,” shares Lin.

Restricting personalised news may in fact increase polarisation under certain conditions on extreme voter’s beliefs and preference. This is why prudence should be exercised, and our research considered, when evaluating the overall impact of personalisation.

Furthermore, Lin’s research has investigated the potential effects of perfect competition among infomediaries, suggesting it could potentially help counter policy polarisation.

 “Overall, our study not only highlights the complex effects of personalised news on the political realm, but provides a fresh perspective on policy implications for democratic processes around the world,” she says.

The College is always keen to explore research collaborations with the public and private sector and to reconnect with alumni. Please get in touch if you would like to know more about partnering with us.