Category Archives: Journalism

The Atlantic vs. Facts

I’m sympathetic to how hard it is for journalists to cover complicated policy areas.

Translating policy into readable articles that both inform and entertain is difficult. In this sense, writing a policy a blog is a luxury. For the most part, I just get to stick to the research and ideas.

But journalists do have a duty to accurately cover the basic facts of the issue they are covering.

In her recent article on Japan’s education system, Alana Semuels included a throw away line about New Orleans education reforms; she writes:

The equity in Iitate stands in stark contrast to a place like New Orleans, which was also hit by a disaster. While Japan’s national government tried to ensure that students in the affected area got more resources after the accident, officials in New Orleans disinvested in the public educational system in their city. Public-school teachers were put on leave and dismissed, many students disappeared from schools’ rolls, and the New Orleans system now consists almost entirely of charter schools.

If you click on the links, which presumably justify her claims, you don’t get taken to supporting research; instead, you get taken to an op-ed.

As it happens, rigorous research has been done on the New Orleans reforms.

In terms of student achievement, after Katrina New Orleans saw some of the biggest gains ever recorded in an urban school district. You can read about the research here.

In the article, the authors of the study note:

For New Orleans, the news on average student outcomes is quite positive by just about any measure. The reforms seem to have moved the average student up by 0.2 to 0.4 standard deviations and boosted rates of high school graduation and college entry. We are not aware of any other districts that have made such large improvements in such a short time.

Rigorous research has also been done on how public spending in education changed in New Orleans after Katrina. Contrary to Semuels claim, spending went up.

The authors of the study write:

New Orleans’ publicly funded schools spent 13% ($1,358 per student) more per pupil on operating expenditures than the comparison group after the reforms, even though the comparison group had nearly identical spending before the reforms.

A quick google search could have turned up both of these studies.

I wish Semuels had taken the time to review these studies.

Because she did not, thousands of readers have been misinformed.

Morality Tales vs. Data Tales


I recently wrote that I think the Upshot (a data-driven, policy endeavor of the New York Times) is at least 5x better than most New York Times articles. I also think is better than most other left leaning publications. Neither are perfect, of course, but it seems like a positive step for journalism as a whole.

The general recipe for their best pieces is this:

  1. Strong data analysis of an issue.
  2. A (fairly) evenhanded description of the pros and cons of the relevant policy choices.
  3. A liberal leaning human story that supports the more liberal policy option.

Ideally, I’d love for the human story to be a little more neutral in which direction it pushes the reader; on the other hand, fair enough; these are liberal publications with generally thoughtful viewpoints, and it’s not crazy for them to use stories that bring these viewpoints alive.

Interestingly enough, I’m not aware of a conservative journalistic project (with a similar reach) that constructs its articles in this manner. This seems to be a pretty big oversight.


In contrast to data drive journalism, too much of regular journalism consists of morality tales dressed up as facts. To use Kling’s three axes model, conservative journalists tell morality tales where the threat of civilization going to hell dominates the narrative. Liberal journalists tell morality tales where threat of someone being oppressed dominates the narrative. And libertarian journalists tell morality tells where government is mucking everything up.

For me, this tendency was probability best seen in the various accounts of the banking crisis, where different journalists either blamed the finance industry (liberal narrative), blamed the homebuyer (conservative narrative), or blamed the government (libertarian / conservative narrative).

I’m not an expert here, but I found the dissenting report by three members of the Financial Crisis Inquiry Commission to be good writing in that it elucidated all the various causes that led to the crisis. The report details how bad actions from the finance industry, homebuyers, and the government, coupled with international capital surpluses and animal spirits – all led to the financial crisis.

Even if you disagree with their recommendations, they at least took a lot of effort to tease out all the various causes. Moreover, regardless of how you weight these causes, if you’re solution set fails to acknowledge and address all causal factors, your remedy will be incomplete, which will lead to more suffering.

Unfortunately, most morality tale journalism: (1) fails to surface all the root causes of the problem (2) fails to elucidate the potential policy options and (3) simply tells a human story that aligns with the tale the reporter wants to tell.


Here’s a potential solution: someone should create a Media Turing Test software program. The software could analyze the article and score it on the three axes. Editors could reject any pieces that scored too far along an individual axis.

Here’s a better solution: stop paying for biased media. If there’s no demand for it, media companies will not deliver it.

Unfortunately, I hold little hope that demand will dry up anytime soon.

What Can We Learn from Stories? A Lot if They are Congruent with the Data


Tyler Cowen once said this of stories:

“As a simple rule of thumb, just imagine every time you’re telling a good vs. evil story, you’re basically lowering your IQ by ten points or more.”

I’m sympathetic to this interpretation.

Stories, like status, often make us dumber.

That being said, I find stories to be very useful in elucidating the emotional components of policy.

Often times, even if I know all the data, I don’t truly “get it” until I hear a story that makes the data come to life.

Now the key here – and I wish every journalist at every paper in the world would get this – is that the best stories follow a fact pattern that the data suggests is common.

Far, far too often, journalists do the opposite. They tell stories as a counterweight to the data.

This consistently happens when reporters cover New Orleans education reform.

Journalists will write something such as: “New Orleans has seen unprecedented gains in student achievement” and then tell a story of a student who is struggling.

This risks leaving readers with the exact wrong impression, as the story is inevitably mentally stickier than the data.

So here’s my plea to journalists.

Tell us great stories.

Just make sure they are aligned with, and not inapposite to, the data.