Photograph by Felbert+Eickenberg/Getty Images
Photograph by Felbert+Eickenberg/Getty Images

A month or so ago, I was chatting with Steve Teles, a political scientist at Johns Hopkins University, about a class he teaches on policy failures. Soon the two of us had conceived a mad plan: I would take his class and blog about it. Because I love class more than almost anything, this would be a great way to combine two of my greatest passions -- and to tell the world about the insights of the brilliant Steve Teles.

To kick things off, I did an instant-message interview with Steve -- sorry, Professor Teles -- about why he teaches the class and what he’s learned from it. Below is a lightly edited transcript.

Megan McArdle: So, first question: What is this class about?

Steve Teles: The class is about large-scale, negative policy consequences -- that is, what explains why policies sometimes work very, very differently than their authors intended. So that's what it's "about" in a topical sense, but it's also "about" developing a particular set of skills in the students -- a "catastrophic imagination" in the students -- an ability to imagine all the things that can go wrong, even about measures that one supports. And also to be able to write clearly and concisely to a principal (a boss) about them.

MM: Why did you decide to teach the class?

ST: I first taught this class in the policy school at the University of Maryland when I was there in 2007-2008. In a professional school, you're really forced to think about what the skills you're adding to the students' portfolio. They're all taking economics, statistics, etc. … so what is it that a political scientist has to add to the training of a master's in public policy?

I thought that reading about big policy mistakes of the past might help these master's students develop what my old friend Marty Levin from Brandeis University calls a "dirty mind" -- an ability to see around corners, which is what sometimes economist-trained people can't, especially about the operation of human organizations.

The other part of it was the Iraq war. Not that anyone cared at the time, but I had supported the war, and yet the whole thing was such a mess, I wondered how I had not applied some of my basic skepticism (which I learned at the feet of conservatives like Martha Derthick, and from reading the Public Interest) to the invasion.

MM: Yes, I wondered the same thing about myself.

ST: So I thought by putting the war in a larger historical context, I'd be able to make sense of it. And so that's why, the first time I taught it, Iraq was all the way at the end. I had students at Maryland who were in the military, so this was not abstract to them. This was business.

So anyways, when I left for Johns Hopkins in 2008, I took the class with me, and have gradually adjusted it every year. Sadly, there's always new material.

MM: Such as?

ST: Hurricane Katrina wasn't in the class originally. Neither was Fukushima.

A few things have always been cornerstones of the course: James C. Scott's "Seeing Like a State,"Diane Vaughan's "The Challenger Launch Decision,"Amy Zegart's "Spying Blind."

But some other cases have come in and out, largely because new policy disasters happened and I had to make room. So this year the last week is on the screwups in the health-care exchanges.

MM: What have you dropped over the years?

ST: So some things have come in and out, but I used to have an interesting book on the tainted blood scandals, when thousands of people worldwide were infected with HIV through blood transfusions. It's come in and out, but the biggest problem is there are like 10 different national cases, and so getting the students to be able to follow what was happening in Japan, France, the U.S., Canada, etc., was just very hard.

I've used Rick Hess' "Spinning Wheels," on the recurrent failure of urban school reform, a book I adore, but that one really was just having to make room. I used to use Mark Rom's book on the savings and loan disaster.

One thing that is a challenge with this course is that it is deeply comparative. We are jumping around across national cases and across time and different policy areas. So it's very challenging in that you have to get on top of completely different basic facts each week. We read David Butler, Andrew Adonis and Tony Travers's incredible book on the poll tax disaster in the U.K. (which is what in the end brought down Margaret Thatcher) … they have to quickly understand the structure of the U.K. political system, and the intricacies of British local government finance, and then the next week they're having to make sense of the Japanese political system and the details of how nuclear plants work. Then they have to understand how a space shuttle operates and the bureaucratic organization of the National Aeronautics and Space Administration.

MM: So what do you want the students to take away from the class?

ST: First, I want to reshape their instincts … I want them to just instinctually look for the things that can go wrong and equip them with a set of cases and mechanisms for failure that are right at the front of their brain.

Second, I want to help them reason from history. Policy makers reason from analogy all the time -- it is one of the most fundamental ways they make decisions. But they do so sloppily, opportunistically, and without a very rich set of analogies.

Third, I want them to be able to communicate, in that form of analogical reasoning, to a boss who doesn't know anything.

Mainly, though, I want them to be able to apply skepticism back at themselves. A big theme in the class is that policy mistakes are caused by selective information processing, our tendency to filter out information that is uncomfortable to our beliefs … I'm trying to equip them with some mental habits to apply skepticism MOST to things they want to believe will work.

MM: I feel like most of us are pretty good at figuring out why policies we don't agree with won't work. The problem is applying that skepticism to your own side, presumably. Is that right?

ST: Well, you'd think so, but no. I mean, it certainly helps to be motivated to find things that will go wrong. But in predicting the mechanism that will cause failure, being motivated doesn't always help. That's why you need a rich set of mechanisms extracted from past experience to know where to look.

MM: It's interesting with the financial crisis and the Iraq war. The people who "predicted the crisis," or said the war was a bad idea, were, by and large, not correct about what happened, or why it was a bad idea.

ST: Yes, that's true. Although, just to be clear, in almost all cases of major policy mistakes, there were people who predicted what would happen. That is, the information that could have allowed you to know what was going to happen was available, but policy makers ignored it or discounted it.

MM: So last question: What have you learned from teaching this class? What has surprised you most?

ST: I think mainly I've learned just how hard it is to reason from history, even though we do it all the time. We all say that the "lessons" of X or Y are whatever, but a "lesson" involves extracting something from one case and applying it to a very different one. That's hard, and easy to do very badly, with terrible effects. So I think the main thing I've learned is a bit more modesty … we have no choice but to reason analogically, but we need to apply a lot of skepticism back on our own reasoning.

I think it made me open to something our buddy Tyler Cowen said, about applying a high level of skepticism to all of our existing beliefs, to think of them all as having some probability of being wrong. That's very hard for normal people like you and me, as opposed to people like Tyler, but it's a good aspiration, and I think ultimately it's the purpose of a liberal arts education, so even as I think this class has a lot of "professional" value, ultimately its purpose is the purpose of all liberal arts education -- to make our thought conscious of itself, to be able to reflect back on ourselves thinking, from the outside.

(Megan McArdle writes about economics, business and public policy for Bloomberg View. Follow her on Twitter at @asymmetricinfo.)

To contact the writer of this article: Megan McArdle at

To contact the editor responsible for this article: Brooke Sample at