The Hype Machine review 5 stars

‘‘What we don’t know is whether these [misinformation] efforts tipped the 2016 election, or if they are affecting the current election in 2020. We don’t know because we aren’t measuring it. To harden our democracies to the threat of digital manipulation we need research and legislation: research to understand the threat, and legislation to neutralize it. Today the threat remains unchecked.’

No great business book can be summarized in 1 sentence. If it can, and thousands each year are, it simply does not do justice to a complex issue. The Hype Machine is a great book, discussing how social media disrupts our elections, our economy and our health, and even offering some solutions! Sinan Aral covers a wide variety of related topics, including what social media does to our brain, and how its hyper-socialization boosts human biases. He describes how local network effects make social platforms grow so fast, how its economic incentives work and how advertising on these platforms is not as effective as many believe, given strong self-selection effects and a lack of causal evidence. Even microtargeting (remember Cambridge Analytica?) is not effective without optimization algorithms.

Evidence and expertise get star roles in Sinan’s book. His references are precise and from different disciplines, and he often relates personal conversations with fellow experts. Moreover, his story-telling is relevant and to-the-point. My readers will recognize this a pet peeve: far too many books waste our time with long rants and unimportant details. Not this book: Sinan simply has too many important insights to share and discuss in a mere 320 pages.

Closest to my current research is Sinan’s discussion of the role of social media in elections. Based on large body of research, we know many voters were exposed to fake news in the 2016 US presidential election, and that it was targeted to specific voter segments by Russian actors. But while ‘the 2019 twin reports of the Senate intelligence committee highlight this, they do not answer, perhaps the most important question facing democracy in the digital age: to what extent are democratic elections vulnerable to social media manipulation?’ I fully agree that is key, and answer it with a resounding yes in our award-winning 2020 paper on how social media drove the 2016 and 2020 elections. The nuance is that the platform and topic matter a lot: while fake news on Hillary Clinton’s emails was responsible for her misfortune in the last 2 weeks of the 2016 election, fake news about her advocating Sharia Law was not. Likewise in 2020, fake news on Biden’s stance on Black Lives Matter and COVID-19 policies reduced his election chances, but other topics, such as alleged fraud and implicating emails, failed to do so. As to the topic of polarization, Boxell, Gentzkow and Shapiro showed it increased steadily over 4 decades for the US, Canada and Switzerland, but not for Germany, Sweden, Australia and New Zealand. I agree that, while social media plays a role in polarization, it is far from the only driver: in the US, it mostly serves to make the income inequality, despair and racial divisions more visible.

Hope and balance are key messages of the book: Sinan goes into detail on the benefits and the perils of social media, believing in the power of our choices to guide us towards the light. His stern warning against breaking up has several good arguments, but some struck me as odd: if e.g. Facebook and Instagram are not competitors (p. 293), why did Facebook felt Instagram was a threat? I do like the 4 levers Sinan identifies (laws, software codes, norms and monetary incentives) and his advice to pull all 4 of them. True data portability and interoperability means messages from one network would be instantly rerouted to the other networks of the consumer. New competitors should get a boost by requiring incumbents like Facebook to provide access to unbundled element of their data-processing infrastructure. Sinan also calls for a national commission on technology and democracy, staffed with bipartisan experts who understand the issues and how they relate.

As to fake news, Sinan recommends labeling (as Facebook and Twitter did in the 2020 election), demonetizing fake news (as YouTube did), stimulating consumer media literacy, and technological solutions and platform policies to foresee and limit the spread of fake news. And how about platform redesign to stimulate good behavior: ‘Instead of the like button, what if there was a truth button, or a trust button or a knowledge button? What if users gained a reputation not through popularity but as being influencers who connected us to the most valuable new ideas and people or who taught us something new, or who provided us with most social support or who corrected our mistakes or saved us from our own bad habits?’ (p.320).

Will these recommendations work? I am hopeful based on my own work: despite the reluctance of Facebook and Twitter to maintain truth in political ads (as described in the book), the 2020 election saw them fact checking and labeling misinformation. While we did see a ‘rerun of the Hillary Clinton email scandal emerge as a Joe Biden Burisma scandal in the fall of 2020’, our analysis finds that ‘but his emails’ was ineffective in driving voter choices, while ‘but her emails’ was responsible for a third of poll variation over the 4 months of the 2016 election.

5/5 Accept as is. Buy this book and let Sinan guide you on your journey!

Leave your comment here

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s