Schrodinger’s Audience: How News Analytics Handed America Trump

Trump’s media dominance isn’t just driven by our attention, it’s driven by the media industry’s new tools for measuring and responding to that attention.


Donald J. Trump is the presumptive Republican nominee. For the Presidency. Of the United States of America.

This is, in fact, happening. We’ve had months to adjust to this alternate reality, in which everything we thought we knew about how American presidential elections work is turned on its head. But today it’s worth pausing to note just how weird it is, and wonder how we got here.

For those who have comfortably adjusted their expectations, please take a moment and imagine having a conversation with the version of you living in May 6, 2015. It was a simpler time. Trump had not yet announced his candidacy. The idea of a Trump presidency was a Simpsons punchline, a gift for Jon Stewart’s farewell tour. As a political scientist, I made light of Trump’s early rise in the polls. We knew what this was. John Sides and Lynn Vavreck refer to it as the “discovery-scrutiny-decline” cycle. In 2012, candidates like Michele Bachmann and Herman Cain had briefly stood atop the polls as well. But media attention leads to public scrutiny, which eventually leads to a polling decline for the non-“serious” candidates. Media coverage and poll numbers rise and fall. It is as predictable as the tides. Candidates like Donald Trump don’t maintain their media dominance for the entire primary season. Their misstatements and lack of party support eventually prove to be their undoing. Last summer, I routinely joked that Trump becoming the nominee would be the equivalent for American political scientists of the collapse of the Soviet Union for the Kremlinologists in the 1990s. (That joke has aged poorly.)

Why didn’t the discovery-scrutiny-decline cycle happen to Trump? It happened to Ben Carson and Carly Fiorina (remember them?). What is it that made Trump different?

There isn’t one singular explanation, of course. Trump is a unique celebrity candidate, with a particular talent for turning subtext into text. And he benefited from a crowded field that featured a never-ending series of strategic pratfalls by his opponents.

But, to my eyes, one feature stands out above the rest: Trump’s utter dominance of media coverage (print, online, and television). Trump didn’t have a brief moment in the sun, followed by a return to normalcy as the media directed attention at his competitors. Even when Ben Carson or Marco Rubio were rising in the polls, Trump was never out of the public spotlight for longer than a moment.

Six weeks ago, Nick Confessore and Karen Yourish covered this point for the New York Times Upshot blog. Donald Trump received approximately six times as much media attention as his closest rival, Ted Cruz. Cruz, you might recall, is also a bit of a showman with a penchant for making wild, offensive claims. (Remember that time he shut down the entire government in what was basically a publicity stunt?)

The simple explanation for Trump’s media advantage is: “Look at the ratings! We’re giving the people what they want.” And that explanation is correct. Trump news, Trump articles, Trump hot-takes all did far better than their competitors. The Republican primaries with Trump have been far more entertaining than the primaries without Trump. CBS chairman Les Moonves has remarked that Trump’s candidacy “may not be good for America, but it’s damn good for CBS.” Ross Douthat has likewise noted that Trump is “such a gift to our industry.”

But I think the simple explanation puts the emphasis in the wrong place. Trump’s media dominance isn’t caused by public demand for Trump. It is caused by the media’s new tools for looking at the ratings!

Gawker CEO Nick Denton once proclaimed that “Probably the biggest change in internet media isn’t the immediacy of it, or the low costs, but the measurability.” Today, all online media outlets rely on sophisticated analytics to track, monitor, and judge which stories are the most popular. Some of these analytics are made public, through “most read,” “most emailed,” “trending” or “related” sidebars on their sites. Others are maintained internally, as a tool to aid editorial judgment.

The thing that we have to remember here is that the act of measuring a process will also change that process. Analytics sends a signal to journalists and their editors (Trump brings in way more traffic than Bush/Cruz/Walker/Clinton/Sanders/Carson). And this signal then helps guide their news routines (people seem to want more Trump, let’s give it to them), which in turn sets up a positive feedback loop, in which Trump’s endless media dominance feeds into the narrative that this election is All About Trump.

In a world with digital media, but less analytics, this election drama would have unfolded differently. Trump would still have been the same showman. He would still have the same instincts, and the same Twitter skills, and the same press conference behavior. But journalists and their editors would have been less attuned to the immediate feedback of Trump’s daily ratings effects, and this would have led them to spread their coverage more evenly (as they always have in the past). Trump’s media dominance isn’t just driven by our attention, it’s driven by the media industry’s new tools for measuring and responding to that attention.

So today, as we think about the all-too-real possibility of a Trump presidency, I think it’s useful to recall Schrodinger’s Cat. Outside the realm of quantum mechanics, Schrodinger’s Cat is usually used as a reference to how the act of observation can affect reality.

We’ve now become Schrodinger’s Audience.

The act of observation has helped to create this reality.