Book Report 11
July 5th, 2025
Freakonomics
Steven Levitt with Stephen Dubner
Why I chose to read this book: I wanted to close the circle. I have read several essays from Freakonomics, along with other essays and also lisetened to the Freakonomics Podcast. While I don’t agree with many points Mr. Levitt makes or his basis for making the points, he is always a provacateur for deeper analytical thinking. There needs to be more of that type of analysis in today’s era of bountiful data.
What I learned from this book: As I mentioned above, I am familiar with Mr. Levitt’s work, but having a series of essays on topics in aggregate, helps form a larger picture of how he systemically approaches his data collection and analysis. The book is divided into 3 sections, and Mr.. Levitt takes the contrarian view in each. The first is on the nature of expertise, the second is on the stereotypes and roots of criminals, and the last is on modern parenting. He is strongest, when his data is strongest. In the case of these three categories, that would be the section on crime. He provides a convincing, if somewhat jarring argument in the 2 essays on the common misconception of criminal success (drug dealers in this case) and the reason crime has abated over the past 30 years (Although his not distinguishing violent from non-violent offenders weakens his arguments a bit in lieu of what has happened over the past 20 years of the surveillance state and the erosion of privacy). He is less successful in his take-down of experts, mainly because the way he chooses to define such people are infields that most us ‘real experts’ know exist due to information asymmetry. In other words, take away the asymmetry, and the ‘expertise’ largely goes away. That being said, the narrow reading of his work here is fairly accurate. His section on parenting is not great, and an unfortunate way to end the book.
What I will take away from this book: Mr. Levitt’s pursuit of data as a way to look into problem we face as a society, and his understanding of how incentives motivate all of us is a lesson we are all better off learning. As I have advanced in my field, and as I look to do more in the future with AI powered solutions, I find myself increasingly relying on massive amounts of data to quantify and verify my work. There are potential answers in the data we all have access to, if we know what questions to ask and how to look at the information. The key, is to try to divorce ourselves from pre-conceived notions of what we might find. Mr. Levitt also provides several methodologies for looking at data that will be helpful to keep in mind in my future endeavors. All in all, Frakonomics is a book well worth reading for anyone looking to broaden their outlook.
