Americans, dazed by the result of how we ended up with the two least likable major-party nominees, may be asking themselves how we ended up here. A new study from Harvard’s Shorenstein Center puts the blame squarely on the media, and not just for Donald Trump’s success. Thomas Patterson’s analysis of the “invisible primary” — the year before voters first got a chance to choose — concludes that the media was more interested in sound bites and “metanarratives” than in sober coverage and substantive debate, and that Trump turned out to be the “mother lode.”

So, although a degree of success in Iowa and New Hampshire is important, neither contest by itself is predictive of the nominee. The better indicator of who will win the nomination is how well the candidates position themselves in the year leading up to the Iowa caucus. This period—“the invisible primary”—is when the candidates try to put in place the ingredients of a winning campaign.

Of all the indicators of success in the invisible primary, media exposure is arguably the most important. Media exposure is essential if a candidate is to rise in the polls. Absent a high poll standing, or upward momentum, it’s difficult for a candidate to raise money, win endorsements, or even secure a spot in the pre-primary debates. Some political scientists offer a different assessment of the invisible primary, arguing that high-level endorsements are the key to early success.^[1] That’s been true in some cases, but endorsements tend to be a trailing indicator, the result of a calculated judgment by top party leaders of a candidate’s viability. Other analysts have placed money at the top.^[2] Money is clearly important but its real value comes later in the process, when the campaign moves to Super Tuesday and the other multi-state contests where ad buys and field organization become critical. …

So what explains the news media’s early fascination with Trump? The answer is that journalists were behaving in their normal way. Although journalists play a political brokering role in presidential primaries, their decisions are driven by news values rather than political values.  Journalists are attracted to the new, the unusual, the sensational—the type of story material that will catch and hold an audience’s attention. Trump fit that need as no other candidate in recent memory. Trump is arguably the first bona fide media-created presidential nominee. Although he subsequently tapped a political nerve, journalists fueled his launch.

Journalists seemed unmindful that they and not the electorate were Trump’s first audience. Trump exploited their lust for riveting stories. He didn’t have any other option. He had no constituency base and no claim to presidential credentials. If Trump had possessed them, his strategy could have been political suicide, which is what the press predicted as they showcased his tirades. Trump couldn’t compete with the likes of Ted Cruz, Marco Rubio, or Jeb Bush on the basis of his political standing or following. The politics of outrage was his edge, and the press became his dependable if unwitting ally.

But he wasn’t the only mother lode. Patterson concludes that Bernie Sanders was a media creation, too, emphases mine:

Sanders’ initial poll position meant that, when he was reported in the news, the coverage was sure to have a negative component. He was in the unenviable position of a “likely loser.” At the same time, his initial poll standing proved advantageous as the year unfolded. As his poll numbers ticked upward, he was portrayed as a “gaining ground” candidate, a favorable storyline buttressed by reports of increasingly large crowds and enthusiastic followers. “The overflow crowds Sanders has been drawing in Iowa and New Hampshire,” said USA Today, “are signs that there is ‘a real hunger’ for a substantive discussion about Americans’ economic anxieties . . . .”  The “real hunger” extended also to journalists, who are drawn to a candidate who begins to make headway against an odds-on favorite. It’s a David vs. Goliath story, the same story that helped propel Gary Hart’s challenge to Walter Mondale in 1984 and John McCain’s challenge to George W. Bush in 2000. A challenger also gives journalists what they relish most—a competitive race. “Hillary Clinton can’t afford to ignore Bernie Sanders any longer,” said a CNN piece. “She has a serious problem on her hands. Sanders is showing that his campaign poses a genuine threat. He is drawing massive crowds months before the caucuses and primaries begin and without much of a staff to speak of.”^[27]

Sanders’ media coverage during the pre-primary period was a sore spot with his followers, who complained the media was biased against his candidacy. In relative terms at least, their complaint lacks substance. Among candidates in recent decades who entered the campaign with no money, no organization, and no national following, Sanders fared better than nearly all of them. Sanders’ initial low poll numbers marked him as less newsworthy than Clinton but, as he gained strength, the news tilted in his favor. On the eve of the Iowa caucuses, Sanders had achieved what was unthinkable in early 2015.  He was positioned to make a credible run at the Democratic nomination.

Chris Cillizza writes that the data changed his mind on the media’s culpability:

I’ve written repeatedly — and self-righteously — about my belief thatascribing the rise of Donald Trump in the Republican primary race to media complicity is ridiculous. And I believed every word.

But, a new study by Shorenstein Center on Media, Politics and Public Policy at Harvard University casts serious doubts on my position as it documents not only the outsized coverage Trump received — from TV and digital media — in the early days of his campaign but also how overwhelmingly positive that coverage was.

Hold the phone, the Atlantic’s Molly Ball objects — it’s the voters that made these choices, not the media:

The Shorenstein report, which assessed coverage by three networks and five newspapers over the course of 2015, is remarkable for its unsubstantiated assertions. Nowhere in the report does its author, Thomas E. Patterson, even attempt to prove that media coverage correlates with candidate standing. He begins by talking about the “invisible primary,” the period of an election when donors and insiders evaluate the candidates before voters begin to tune in. “Of all the indicators of success in the invisible primary, media exposure is arguably the most important,” Patterson writes. “Media exposure is essential if a candidate is to rise in the polls. Absent a high poll standing, or upward momentum, it’s difficult for a candidate to raise money, win endorsements, or even secure a spot in the pre-primary debates.” The word “arguably” is a good tip-off here: It signifies an absence of evidence. There are no footnotes or other references for this string of assertions, simply the author’s contention that they are true. …

Patterson goes on to cite data that finds that Trump enjoyed more “positive or neutral” news coverage than the other Republican primary candidates, and that the volume and tenor of early coverage were disproportionate to voters’ interest in Trump’s candidacy. But when you look at the report’s evidence, the preponderance of this coverage turns out to be reporting on the fact that Trump was gaining in the polls and attracting large crowds to his events. The press is the effect here, not the cause: The media were noting—often to their collective surprise—that more and more Republican primary voters were becoming receptive to Trump’s message. Should they instead have ignored or downplayed this development? Journalists shouldn’t blindly follow polls, but we should—constantly!—attempt to understand what sentiments are percolating in the electorate.

Hmmm. Both raise good points … which is why my column for The Fiscal Times spreads the responsibility to the politicians, voters, and the media:

In other words, once Sanders becomes the sincere David to Clinton’s calculating Goliath, the media reinforces that metanarrative in all of the coverage that follows. The same holds for Trump’s position as the man with his finger on the pulse of working-class America, a rather ironic position for a billionaire real-estate mogul who has contributed vast amounts of money to both parties. Patterson doesn’t mention the metanarrative about Hillary Clinton and glass ceilings, but it’s also been a constant metanarrative in this cycle. In a sound-bite coverage environment, the incentives line up towards playing up the metanarratives rather than explore ways in which they may not apply. …

The irony of all this is that the media still matters. Polling consistently ranks news media near the bottom of all public institutions, yet they apparently retain enough influence with voters to turn one major-party primary race and nearly turn another. However much the media exploited Trump for ratings and Sanders for narrative, the truth is that voters responded to it, according to Patterson and Shorenstein. We bought the media’s creations.

For representative democracy to work properly, it requires a well-informed electorate and responsible journalists to deliver the critical information needed for that success. When both politicians and journalists become more concerned with entertainment than illumination and voters hold neither accountable for it, we should not be surprised when we end up with caricatures as our final options on Election Day.

Call that the Murder on the Orient Express conclusion — everyone did it. But even that might not encompass the entire scope of responsibility. Glenn Reynolds puts the ultimate blame on the office itself, which has grown too far in stature and has become a celebrity platform all its own:

The presidency as it exists today is a mess. Presidents have too much power, too little accountability and too high a public profile. That makes the job attract the wrong sort of people, and then ensures they’re not up to it.

If we were to shrink the government, and shrink the presidency, we might find that what was left would attract better people — and would be easier even for lesser mortals to execute.

When your political system consistently delivers bad results, it’s time to look at a change.