November is in the books, so it’s time for an update to our Fiat News Index. A few refreshers from prior iterations:
- What we call Fiat News is not the same as bias. You’re welcome to take charts and exhibits from this In Brief, but we’d be grateful if you’d keep this context in mind. For a primer on what we mean by Fiat News, please take a look at the original piece Ben wrote in 2017 here.
- FNI looks through articles for the presence of 10 words we have identified as having a strong relationship with explanations, assumptions or implicit claims of causal links, or words asserting common knowledge or self-evident value judgments.
- Our aim is to provide a rough estimate of how often an outlet’s articles are explaining to readers how to think about a topic.
- We have received a lot of good feedback on expanded or modified versions of these rules – thank you! We want to be deliberate in exploring these ideas. We also want to avoid changing too many things in the framework at once. This go around, we have kept the existing rules, and instead emphasized some requests to explore how the tendency to use fiat news words changes for different topics, which we show below. We will be exploring the ruleset more in future editions.
- We have expanded our universe of news providers to 30, including all of the top sites used as sources on social media (except Buzzfeed – sorry, not sorry). We are aware, and you should be aware, that these sources are not all pure news sites. Many are commentary and analysis sites whose stated aim is to explain or provide opinions on topics. For this reason, we aren’t suggesting that a high index score should always be viewed pejoratively; for example, we named our unit of fiat news the “Vox”, as in, “This site does about 0.8 Voxes of explaining the news to you,” but that’s because that’s literally what Vox says they do. Far be it from us to add to the chorus of people telling you how to think, but we are more concerned about outlets purporting to be sources of fact which demonstrate high indexes. The commentary/analysis websites simply provide a good point of comparison, or (if necessary) a reminder that they aren’t a substitute for news for the independent-minded.
- You should be aware that two large publications – The Wall Street Journal and Financial Times – remain outside of our dataset, which is a Moreover- and LexisNexis-powered database accessed through Quid.
Fiat News Index for the Month Ended November 30, 2018
As noted above, for this month we further divided our universe into three sub-topics: Trump, Markets and Climate Change. The Trump sub-topic is defined simply as articles with references to “Trump.” The Markets sub-topic is defined as articles referring to “Markets”, “Stock Market(s)” or “Stocks.” The Climate Change sub-topic is defined as articles referring to “Climate” or “Warming.” Each of these words will pick up some false positives, but that effect appears to be limited, and our feeling was that the work stands up better without manual overrides. In any case, the data is based on the rate of fiat news terms and not the total number of articles about a topic. Please note that we have excluded outlets with fewer than 10 FNI-containing articles relating to one of these sub-topics to avoid over- or understating their tendencies.
“Trump” Articles FNI for the Month Ended November 30, 2018
“Markets” Articles FNI for the Month Ended November 30, 2018
“Climate Change” Articles FNI for the Month Ended November 30, 2018
Sub-Topic Differences vs. Normal
While it may not be immediately apparent from the way the data is presented in isolation, you won’t be surprised to learn that the general tendency to use explainer words is much higher on each of these three topics. The chart below illustrates this on an outlet-by-outlet basis. The differences are multiplicative, not additive, and can be read as, “How much more likely am I when reading a story about [Topic] by [Publication] to have the topic explained to me than in a typical article by [Publication]?“
Without considering the question of persistent biases, which is a much more complicated question, what are some early takeaways here for the citizen who wants to receive as little in the way of story-telling, analysis or explanation of their news as possible?
- Going straight to the source with Reuters and Bloomberg continues to looks like a good choice.
- Whether because of reliance on these and other wire services, or because of its own newsroom standards, the SF Chronicle continues to stand out among print options.
- If you are reading Fox News, the BBC, NY Times or Washington Post for political news, that news was explained to you more often in November than it would have been in almost any other major news outlet – at levels comparable to those of pure commentary websites.
- Beware the paradox of science journalism: if climate science coverage over the last month is in any way indicative, it is filled with more explainers and a lower percentage of just-the-facts-ma’am coverage than even some of the most polarizing political topics.
Our plan for December is to explore and continue to refine our rules.
Update: Modified the ordering of the X-Axis of the final chart to match the initial FNI exhibit for ease of comparison.