Well, I am sad to say I wasn’t that impressed. It is definitely written by a journalist rather than by an academic.
Absolutely you should read or listen to one of the author’s many interviews on how he broke his afternoon cookie habit, but once you’ve done that, you really have the substance of the worthwhile portions of the book.
Why do I not like this book that I was predisposed to like? Well, I don’t trust or believe it. He uses a lot of examples that I’ve seen before but that illustrate completely different concepts and were set up to illustrate different concepts and don’t actually fit in the way that he’s saying if you look at the original research (plus he’s got the simplified journalist-reported version of the marshmallow experiment based on something that doesn’t really exist rather than the original studies). Secondly, the narrative often seems more important than the truth. He doesn’t have a good conceptualization of the idea that correlation is not causation and will force causation where it doesn’t actually belong because that is what makes the story look good. I suspect he’s done that in some of his reporting of people’s stories as well– they seem too simple, too uni-directional, to be true.
And he is aware of these flaws– rather than footnotes or traditional endnotes, he has chapters of notes in the back of the book sorted by page number. They’re difficult to connect with the narrative. But anytime I found myself going, “Wait a minute… that’s not right,” I would flip to the back and there would be an admission that no, what he was glibly saying didn’t actually pass fact-checking. The government official in question says that yes, although the infant mortality rate decreased, he cannot take credit for it. (But phrased in a way that makes the official sound humble, which flows with the narrative Duhigg has created about him, rather than as someone who wants to make it clear that correlation is not causation.) There’s a huge debate about whether or not 12-step programs like AA work, with general scholarly opinion finding that they don’t, which he notes in a note. For a reporting story about a hospital, he notes again that he’s compiled different stories and has left out the ones that disagree rather than agree with his narrative (but phrased in a way that makes it sound like those who disagree are lying). Almost every time I thought something was too pat to be true or I knew that the actual research wasn’t so simple, I’d flip to the back and there would be a note admitting that no, it isn’t so simple.
But he didn’t let that interrupt his storylines.
So yes, I think his story about how he broke his cookie habit is useful and compelling. You can find it in the appendix. But the rest of the book, not so much. More disappointing than Malcolm Gladwell. (Interestingly, I overheard one of DH’s audible books by Doris Kearnes Goodwin while reading The Power of Habit and figured out why academic historians have problems with popular historians… so much attribution without evidence. It hurts!) The book is not without value, but many of its stories and conclusions should be taken with a grain of salt.
I do, however, still strongly recommend Willpower by Baumeister and Tierney. Probably the best types of these books are collaborations with academics and reporters. (Or occasionally there will be a great book written by an academic who also happens to be a good popular writer.)