Red in the Face

As front page headlines chronicle the ever-growing daily drumbeat of Obamacare woes, the saturation of domestic coverage has been displacing a wide range of critical foreign policy developments. Syria’s bloody civil war rages on, while Egypt potentially stands on the brink of its own. The United States and Iran, meanwhile, forge historic progress towards a nuclear agreement — not that you’d know it from the headlines, which remain focused on lost health insurance plans, faulty computer code, and sweetheart government contracting deals. Official confirmation of the Syrian government’s destruction of its chemical weapons production facilities barely garnered a mention, while CBS News’ retraction of a 60 Minutes piece on what went wrong in Benghazi — one of the few retractions issued in its 45-year history — yielded a collective yawn. Indeed, had one woken up this morning after a year-long nap, one could be forgiven for assuming that the most pressing U.S. foreign policy issues had magically been resolved, leaving Obamacare as the single greatest threat to domestic tranquility.

While technically a domestic policy issue, the political turmoil that has enveloped Obamacare has had a significant impact on U.S. foreign policy — from sparking the total shutdown of the federal government to forcing the Obama administration into damage control and public outreach at a critical foreign policy juncture. The constant distraction created by the Obamacare debacle, meanwhile, is having an impact of its own. The latest Gallup poll, current through Nov. 13, shows that since the government shutdown, the president’s disapproval rating has reached close to the highest levels of his tenure.

Coupled with the daily drumbeat of NSA disclosures, the Obama administration is finding itself playing near-constant defense. This raises key questions about what the national coverage of Obamacare really looks like, how all-consuming it is, and how the country as a whole is reacting to it. Is the cacophony of headlines foretelling ultimate healthcare doom truly reflective of a darkening across the nation towards Obamacare? Are they really eating up political capital that the administration should be leveraging toward other activities, including foreign policy?

Even in the increasingly online realm of the news media, there is still a fixed “news hole” that can only accommodate a small number of stories. High profile coverage of Obamacare knocks stories about American foreign policy off the front page, shifting the public’s attention back home and decreasing the average citizen’s knowledge of the latest developments abroad. One way to visualize this phenomenon involves using big data, which paints an unusually vivid picture of the pattern and tone of news coverage. Let’s take a look at the numbers and see what we can learn about what the media picture of Obamacare really looks like and how it has evolved.

CHARTING TIME AND TONE

So how much is too much? The graph below shows the percentage of all American television news programs monitored by the Internet Archive’s Television News Archive that mention Obamacare or the Affordable Care Act. Starting in mid-2011, the law has been a constant fixture of television news, with the brief exception of the April 15-28 period, when it was temporarily displaced by round-the-clock coverage of the Boston Marathon bombing. Beginning on Sept. 18, as the GOP moved to tie defunding Obamacare to continued funding of the government, coverage began to skyrocket, reaching 81 percent of all monitored television news by Oct. 1, the first day of the shutdown. Even with the restoration of the government, the postmortem commentary and almost immediate website failures have kept Obamacare a fixture in at least half of all news programs each day throughout the month of October.

Television captures just one dimension of the news environment, so we turn to something called the GDELT Global Knowledge Graph, a massive compilation of the world’s people, organizations, locations, themes, emotions, and events culled from the global news media. Using GDELT data for the period between April 1 and Oct. 31 2013, we get the graph below, which compares the relative volume and “tone” of the discussion of Obamacare. To make the numbers easier to interpret, the graph uses what’s called a “z-score,” which calculates the average of a data series and then for each point reports how many standard deviations it is away from that average. Put another way, a z-score tells us how different each data point is from the rest of the data points. This allows us to plot very different datasets on the same graph to compare how they change over time. (Note for math geeks: the actual z-score is shifted by its min/max so that volume is positive and so that tone positivity/negativity matches the scale). Here tone is computed as how “positive” or “negative” the coverage is — based on the relative density of different emotionally charged words in the text — with higher numbers indicating more positive coverage and lower numbers being more negative.

Instantly, it is clear that the GOP’s tactic of linking Obamacare to the U.S. government shutdown had the intended effect of throwing it into the spotlight, more than doubling its average daily coverage even after the government reopened. The tone of Obamacare coverage seems to have been solidly negative over the past half-year, beginning a sharp descent towards very strong negativity as the shutdown approached and rebounding slighting in its aftermath. As of the end of October, with stories of Obamacare website failures in full swing, tone was still improving — reflecting both counter-coverage of those gaining healthcare for the first time and the fact that even the failed website news was still better than the depths of the government shutdown — but was still strongly negative. 

How does television coverage of Obamacare compare with the broader news ecosystem?

The graph below shows just that, using data from GDELT to compare the intensity of coverage in the online news sphere with coverage on television. The major point to note here is that the online sphere is characterized by a far more gradual ramp-up of Obamacare coverage, peaking on October 1 — the start of the shutdown — and then immediately ramping downward substantially. Television news, meanwhile, shows an almost vertical jump over a short, three-day period to hit its highest level on Sept. 20. It then remains far more elevated after the shutdown than the online news sphere. This suggests that television coverage of major issues tends to fluctuate in much greater extremes, with sudden, massive surges that border on fixation. This is especially consequential because television has a fixed amount of airtime, meaning that increased focus on Obamacare necessarily displaces discussion of other issues, such as Iran and Syria.

How does the tone of online news coverage stack up against that of television news? The graph below looks at this comparison, computing each day as the average tone of the previous three days to smooth out the graph and make the patterns clearer. The obvious takeaway is that the two are highly aligned: Television and online news both largely present the same picture of Obamacare. This is not unexpected
, but is nonetheless important because it suggests that measuring one medium in this case gives us a good picture of the other.  

THE GEOGRAPHY OF DISCOURSE

Perhaps more interesting than asking if the nation is turning toward or against Obamacare is asking what parts of the country are for it and what parts are against it. Is it the case that some areas of the country are still strongly supportive of Obamacare while others are not? Was interest in Obamacare initially focused in key urban areas before spreading across the country as the media frenzy intensified? Are there differences between north and south, east and west, urban and rural? 

There are two approaches that can be used to measure the geography of discourse around a topic. The most intuitive is simply to place each news outlet where it is geographically published (for print) or where it is broadcast (for radio or television) and aggregate views across news outlets based in the same geographic area. In the era of online media and cable television, however, it is difficult to assign a geographic location to outlets like the Huffington Post or CNN. Moreover, the national focus of Obamacare means that while it is covered by outlets across the country, they often focus on discussion of other areas of the country, such as Washington, D.C. or the widely-covered examples of changes in California’s healthcare system. In cases like this, it is more useful to look at which areas are being discussed the most in the context of Obamacare — and what tone is associated with them.

So instead of looking at what is published where, the images below use an approach known as “full text geocoding,” which essentially uses computer algorithms to read through the entire text of a news article, identifying any locations — for instance, “Cairo” — and using the surrounding context of the article to disambiguate whether it refers to Cairo, the capital of Egypt, or Cairo, the small town in Illinois.

The result is a list of locations found in the text and approximate coordinates for each that can be placed onto a map. In addition, the images use a powerful heatmapping tool, specifically designed for media data, to generate a set of unique heatmaps that take into account the total media attention paid to each city, the percent of that attention mentioning Obamacare, and the tone of that coverage. (The software takes into account that a place like New York City garners far more news coverage than a small town like Cairo, Illinois and thus generates a map that normalizes for these factors.) The end result generates an approximation of what the collective news media — regardless of where it is based — is saying about each location in relation to a given topic.

Using this approach, the following sequence of maps shows the national tone toward Obamacare by month, from June through October 2013. Dark red is extremely negative while dark blue is highly positive. The large areas with no color are those that did not receive enough media attention to merit notation, while the small “air gap” in areas with large amounts of color represents the area where highly negative and highly positive coverage meets. Looking at the sequence of maps, one can see how Obamacare spread across the national news consciousness, almost like a weather front, and just how negative the entire nation had become by the end of October, when Obamacare became intertwined with the government shutdown and the new website experienced its catastrophic unveiling.

PUTTING IT ALL TOGETHER

So what does this all mean and what are the policy takeaways? One might reasonably argue that none of the results here are altogether surprising — it doesn’t require data mining to figure out that media coverage of Obamacare has become deeply negative over the past year or that it is consuming a large fraction of the daily headlines. Yet, what we are able to see in the crisp mathematical precision of the computerized graphs and maps above is just how vast and intense the negative coverage really is. As a result, we can move beyond anecdotes like “It’s getting a lot of coverage” to precise statements like “More than 80 percent of all television news shows are talking about it.”

We can also gaze through the eyes of the news media and literally map the deep pessimism towards the law as it spreads across the nation. This by itself is a key finding: just how much the media has been covering Obamacare and, in particular, how key the GOP’s tying of Obamacare to the government shutdown was in bringing it to the forefront. Indeed, while Republicans may have lost in their attempt to defund Obamacare through their shutdown showdown, they succeeded in making it a national news item, and thus setting the stage for the media to eagerly pounce on the first hints of a problem with the new website.

A month after the government shutdown, more than 60 percent of American television news programming still discusses Obamacare, while a vast array of critical foreign policy issues struggle for coverage amongst this deluge. Of course, this is simply what the news media does — across the world, it reports on the freshest stories that are likely to win the most readers. Even with the potentially infinite virtual space of the online world, there is still a fixed amount of real estate on the front page, fixed number of reporters, and a fixed amount of time in the day to cover all the stories competing for attention. Still, the sheer magnitude of the shift inwards caused by the Obamacare debacle and the attendant loss of political capital and public approval have real implications for the administration’s flexibility in tackling future foreign policy issues.

For the first time, we can use sophisticated computer algorithms to transform the daily heartbeat of the news media into a visual window onto the national consciousness. Through the powerful lens of big data we now have a telescope to peer not at the heavens above, but at ourselves here on Earth — revealing a frightening collection of stormclouds that have thrown a shadow over American foreign policy.