The book I most regret having read

13 May, 2026

...is Thomas Piketty's Capital in the Twenty-First Century. I read it, if memory serves, around the summer of 2014, soon after it came out in English.

It's my most-regretted book for simple reasons. The opportunity cost was large (696 pages) and I didn't benefit from it much:

1. I didn't learn much theory from the book: the good mainstream criticisms of the theory seemed to me to have by far the best of the debate.¹
2. Piketty's exposition of his empirical / data-collection work (i) wasn't particularly clear or useful to read in full and (ii) also is not holding up well (as far as I can tell).²
3. More generally, getting Piketty-in-full had little (or negative) value compared to the best summaries and reviews. The book was praised as a landmark book, but in hindsight the blogosophere covered it remarkably well, and that discussion is holding up much better than the book itself. Note that I'm much less likely than most readers to say this about a book!
4. It just hasn't stuck with me very well. I find the experience of having read mysterious and only somewhat predictable; one reason I maintain a list of favorite books is to track how I think about them months and years (or decades) after I read them. When I think back on Capital in the Twenty-First Century, I remember mostly (i) reading r>g over and over, (ii) circumlocution, and (iii) a lot of Père Goriot references. (I was annoyed by how repetitive the Père Goriot references were, but they were still probably my favorite part of the book.)

How did I wind up making this mistake?

1. I was excited about economics at the time, and the conversation about the book excited me.³ It seemed most responsible to actually read the book.
2. To the extent I have political-economic views about these questions, they are... not what Piketty advocates for. I was hoping to find the book broadening and challenging.
3. A friend wanted to read it with me. (Much to his credit, he did in fact read and discuss it.)
4. I was still finishing my Ph.D. and was more in the habit of slogging through academic prose.⁴

Two generalizations: (i) popular social science books do a bit worse than average at offering value beyond their authors' New York Times guest pieces (or whatever) and (ii) I've rarely found reading groups, whatever other virtues they might have, to be a good way to learn from books.

For whatever it's worth: I think that opportunity-cost analyses of books are often tricky and misleading. For example, I could have read a ton of other stuff with the time and effort I put into reading War and Peace cover to cover. (The expected value of (say) As I Lay Dying, a bunch of Shakespeare, and several other books is very high!) It seems indecent, though, to focus too much on value per unit effort, and I'm not just saying this because I like long books. I'm still eager to tackle 696-page books, even if I won't on average consider them twice as good as 348-page books. But this one was a mistake.

#productivity #reading