Thank you for your in depth reviews. So many book reviews tend to be brief, with quite a bit of fluff. Your reviews helped me decide whether I wanted to invest the time in reading some of them, and hopefully avoid the disappointment I often feel when a book is underwhelming.
I'm not able to read Political Epistemology, but it's a topic that has been severely neglected so it's good to see it given more attention.
Curious if the topic of causal explanation comes up. There is a fundamental tension between the idea (in causal modeling circles, and going back to Hume, really) that causation requires invariance in some general relationship. However, all measured relationships between variables in social behavior change over time...sometimes abruptly in response to new conditions. This gets swept under the rug with statistical accounts of causation, or general handwaiving, that don't resolve the tension.
Am I getting my hopes up too much to think that Hannon and Woodard address this?
I had not read this article, thanks. It is a good overview of what I consider the conventional view on causation today, but almost all of the discussion relates to problems created by complexity, not variability. That is, other than a couple examples that aren't really explored, there is no discussion of what it means for the coefficients in a DAG to change over time, across different measurement periods.
The standard view is that causal explanation is hard because the world is super complex. What I'm saying is that the problem is deeper. In physical science there are mathematical laws and models that find invariant relationships. In social science there are not, and as far as we can tell there never will be. So....what does that mean for our use of DAGs and other causal modeling methods? Are we just bullshitting and imputing invariance where we know it does not exist?
Econometrics gets close to what you're talking about. You wrote a paper Moral Goals in Scientific Clothes and raised the thorny fact-value problem. Econometrics, by reducing economics to mathematics, offers to solve the problem, but alas faces the criticism that it is based on political-economic values (currently neoliberal or Monetarist ones, but they could be any values; before the IMF changed its rules in Jamaica 1976, they were Keynesian values).
Yes, I would say econometrics is a prime example of the attempt to put moral goals in scientific clothes, precisely by proposing itself as a "solution" to the problem and then wandering off to the next technical puzzle whenever one brings up that it hasn't.
Great stuff! I think lists like these are very good ways to discover new reads, since recommendation engines are mostly good for ‘more of the same’, and aggregate ratings are too confounded (on Goodreads, the Harry Potter series is higher rated than nearly *every* 20th c. nonfiction book).
I can see that, but I find that “high quality” and intellectual memoir and autobiography contain knowledge that can prove very useful. In fact some of the best books that I read this year were either memoir or autobiography; John Mackey’s “The Whole Story” and Rob Henderson’s “Troubled” being just two.
Certainly Adam Smith’s TMS isn’t memoir, but in a way it contains miniature memoirs, that prove useful. Unfortunately I haven’t read much of it.
Coming from a natural science background, I find Peter Turchin’s economic and social analysis of cycles in historical polities intriguing, so I would encourage you to bump it up your list!
Thank you for pointing me toward’s Yascha Mounk’s critique; I’ll read it with interest
I’ve had the chance to read Yascha Mounk’s critique – thank you again for linking to it.
Having read End Times, I find that Yascha largely misses the point – he focuses solely on elite overproduction in isolation and completely ignores the rest of Peter Turchin’s model, especially his point around “popular immiseration”. Yascha even mentions jaded working class Trump voters in his critique without mentioning that Peter covers this in his book as a central part of his model, which is either careless or disingenuous.
The main point of Peter Turchin’s theory is that the socioeconomic class structure of any given society is a Complex System, and it undergoes similar cycles of stability and instability throughout history (not just in present day America, which is where Yascha focuses his critique).
The core of Peter’s model is the concept of the “wealth pump” — rising inequality leads to increasing political power, wealth, and status for those at the top of society, and worsened conditions for those at the bottom of society. This increased power for those at the top (aka the elite) is self-reinforcing, as economic power captures political power, and has the side-effect of dramatically increasing competition for places at the top of society (this is where elite overproduction enters the model) – as the middle class is hollowed out, people either fall into hardship or make it into the increasingly oversubscribed upper rungs.
Now, it’s perfectly valid that this model should be critiqued, but I strongly disagree with the idea that it is possible to evaluate the strength of a model of a complex system by criticising a single element of it in isolation — elite overproduction (if it is an observable quantity, as Peter claims based upon the large data set collected by Seshat) is both determined by and dependent upon all the other factors governing the structure of society. It would be the equivalent of critiquing the Lotka-Volterra equations by disagreeing with Volterra’a methods of measuring the catch rates of predator fish while ignoring the rest of the dynamics inherent to the model.
To summarise, the core of Peter Turchin’s model is the “wealth pump”, which is a consequence of rising inequality, of which elite overproduction is only one element (even if this is the element which has most caught the public imagination).
Thanks for the reviews. I think you're misreading Ritchie's book though. Apart from creating s straw person by claiming most people believe climate change is an existential risk (most polls don't show that and it's not what most students I talk to believe) thousands of scientific studies show that we're on a trajectory of 2.5-2.9C average temp rise by 2100 and the need to reduce emissions is extremely urgent. Sure, these changes are unlikely to be existential, at least for us, but they will almost certainly cause vast suffering including the destruction and loss of species and ecosystems. Is that alarmist? Even without direct existential threats climate change could lead to them indirectly through geopolitical instability, great power conflict, etc. This is why organisations that evaluate global catastrophic risks often include or mention climate change. But unlike other global threats like advanced AI for which there is considerable uncertainty and disagreement about when and even if they occur, harms from climate change appear virtually certain, even at relatively low levels of warming. So yes, we should celebrate our success as the first generation to (sort of) achieve social sustainability (the core message of the book in my reading) but the stakes are still enormous and the need for immediate concrete action remains. In short I think you're understating climate risks and misrepresenting the books conclusions.
Omfg based and Fukuyama-pilled. My goal is to read everything Fukuyama has written EXCEPT End of History so if I ever meet him I can nerd out about all his stuff while acting completely oblivious to his most famous (and misunderstood) book.
"End of History? Is that your new book or something? I haven't heard about it"
I look forward to your extended reviews on the 1st 2 titles. I haven't read Di Resta yet, but Al-Gharbi was great. I was reading it during the US election in November, and everything seemed clearer to me.
Still, I think I agree that he is excessively woke, and I would like to see how you flesh that out. He reminds me of Veblen in that he takes the stance of a detached, objective socal scientist to deliver a half disguised semi religious jeremiad against the hypocrisy of his class.
On Marx, I recommend the highly readable book by Thomas Sowell, himself an ex-Marxist. Impressively, he presents Marx objectively throughout until unleashing on him in the final chapter. Actually, that's much the way I did it for my Marxism course at Oxford. My tutor liked me until that last class. It would be hard to find a more readable book about Marx than this one.
Surprised you've read a book on Marx in 2024. Is any more clarity required? His Communist Manifesto is pretty clear to me. While the vast majority of Western Academe is heavily influenced by him, many, though still critical of the West, now contradict him, particularly on Materialism versus Idealism. The scandal at your current university over Kathleen Stock epitomises this contradiction.
The Stock outrage is also why the upcoming book on Political Epistemology is of particular interest, so will you be reviewing it before its May 2025 publication?
Thank you for your in depth reviews. So many book reviews tend to be brief, with quite a bit of fluff. Your reviews helped me decide whether I wanted to invest the time in reading some of them, and hopefully avoid the disappointment I often feel when a book is underwhelming.
I'm not able to read Political Epistemology, but it's a topic that has been severely neglected so it's good to see it given more attention.
Curious if the topic of causal explanation comes up. There is a fundamental tension between the idea (in causal modeling circles, and going back to Hume, really) that causation requires invariance in some general relationship. However, all measured relationships between variables in social behavior change over time...sometimes abruptly in response to new conditions. This gets swept under the rug with statistical accounts of causation, or general handwaiving, that don't resolve the tension.
Am I getting my hopes up too much to think that Hannon and Woodard address this?
Looking at the book's Contents, I doubt it, but here's an article you will probably have read and by an author you probably know that addresses the tension you raise > https://www.thenewatlantis.com/publications/correlation-causation-and-confusion
I had not read this article, thanks. It is a good overview of what I consider the conventional view on causation today, but almost all of the discussion relates to problems created by complexity, not variability. That is, other than a couple examples that aren't really explored, there is no discussion of what it means for the coefficients in a DAG to change over time, across different measurement periods.
The standard view is that causal explanation is hard because the world is super complex. What I'm saying is that the problem is deeper. In physical science there are mathematical laws and models that find invariant relationships. In social science there are not, and as far as we can tell there never will be. So....what does that mean for our use of DAGs and other causal modeling methods? Are we just bullshitting and imputing invariance where we know it does not exist?
Econometrics gets close to what you're talking about. You wrote a paper Moral Goals in Scientific Clothes and raised the thorny fact-value problem. Econometrics, by reducing economics to mathematics, offers to solve the problem, but alas faces the criticism that it is based on political-economic values (currently neoliberal or Monetarist ones, but they could be any values; before the IMF changed its rules in Jamaica 1976, they were Keynesian values).
Yes, I would say econometrics is a prime example of the attempt to put moral goals in scientific clothes, precisely by proposing itself as a "solution" to the problem and then wandering off to the next technical puzzle whenever one brings up that it hasn't.
Great stuff! I think lists like these are very good ways to discover new reads, since recommendation engines are mostly good for ‘more of the same’, and aggregate ratings are too confounded (on Goodreads, the Harry Potter series is higher rated than nearly *every* 20th c. nonfiction book).
You got a lot of reading done considering your writing output and presumably your teaching/academic job. Why do you tend to avoid memoirs?
Just don't tend to find they contain useful knowledge I can apply in other contexts - so at best they typically have entertainment value for me.
I can see that, but I find that “high quality” and intellectual memoir and autobiography contain knowledge that can prove very useful. In fact some of the best books that I read this year were either memoir or autobiography; John Mackey’s “The Whole Story” and Rob Henderson’s “Troubled” being just two.
Certainly Adam Smith’s TMS isn’t memoir, but in a way it contains miniature memoirs, that prove useful. Unfortunately I haven’t read much of it.
I am curious about the books that you liked mostly based on entertainment value now
Brilliant, thanks for sharing, Dan.
Have you read Peter Turchin’s End Times (published in 2023)? I’d be interested to know what you think of his analysis
It's on list. I found Yascha Mounk's critique of the idea of elite overproduction quite persuasive: https://www.persuasion.community/p/there-is-no-surplus-elite-in-america
Coming from a natural science background, I find Peter Turchin’s economic and social analysis of cycles in historical polities intriguing, so I would encourage you to bump it up your list!
Thank you for pointing me toward’s Yascha Mounk’s critique; I’ll read it with interest
I’ve had the chance to read Yascha Mounk’s critique – thank you again for linking to it.
Having read End Times, I find that Yascha largely misses the point – he focuses solely on elite overproduction in isolation and completely ignores the rest of Peter Turchin’s model, especially his point around “popular immiseration”. Yascha even mentions jaded working class Trump voters in his critique without mentioning that Peter covers this in his book as a central part of his model, which is either careless or disingenuous.
The main point of Peter Turchin’s theory is that the socioeconomic class structure of any given society is a Complex System, and it undergoes similar cycles of stability and instability throughout history (not just in present day America, which is where Yascha focuses his critique).
The core of Peter’s model is the concept of the “wealth pump” — rising inequality leads to increasing political power, wealth, and status for those at the top of society, and worsened conditions for those at the bottom of society. This increased power for those at the top (aka the elite) is self-reinforcing, as economic power captures political power, and has the side-effect of dramatically increasing competition for places at the top of society (this is where elite overproduction enters the model) – as the middle class is hollowed out, people either fall into hardship or make it into the increasingly oversubscribed upper rungs.
Now, it’s perfectly valid that this model should be critiqued, but I strongly disagree with the idea that it is possible to evaluate the strength of a model of a complex system by criticising a single element of it in isolation — elite overproduction (if it is an observable quantity, as Peter claims based upon the large data set collected by Seshat) is both determined by and dependent upon all the other factors governing the structure of society. It would be the equivalent of critiquing the Lotka-Volterra equations by disagreeing with Volterra’a methods of measuring the catch rates of predator fish while ignoring the rest of the dynamics inherent to the model.
To summarise, the core of Peter Turchin’s model is the “wealth pump”, which is a consequence of rising inequality, of which elite overproduction is only one element (even if this is the element which has most caught the public imagination).
Just began reading The Origins of Political Order!
Thanks for the reviews. I think you're misreading Ritchie's book though. Apart from creating s straw person by claiming most people believe climate change is an existential risk (most polls don't show that and it's not what most students I talk to believe) thousands of scientific studies show that we're on a trajectory of 2.5-2.9C average temp rise by 2100 and the need to reduce emissions is extremely urgent. Sure, these changes are unlikely to be existential, at least for us, but they will almost certainly cause vast suffering including the destruction and loss of species and ecosystems. Is that alarmist? Even without direct existential threats climate change could lead to them indirectly through geopolitical instability, great power conflict, etc. This is why organisations that evaluate global catastrophic risks often include or mention climate change. But unlike other global threats like advanced AI for which there is considerable uncertainty and disagreement about when and even if they occur, harms from climate change appear virtually certain, even at relatively low levels of warming. So yes, we should celebrate our success as the first generation to (sort of) achieve social sustainability (the core message of the book in my reading) but the stakes are still enormous and the need for immediate concrete action remains. In short I think you're understating climate risks and misrepresenting the books conclusions.
Fair enough. Thanks for the comment. I don't think we disagree actually.
Omfg based and Fukuyama-pilled. My goal is to read everything Fukuyama has written EXCEPT End of History so if I ever meet him I can nerd out about all his stuff while acting completely oblivious to his most famous (and misunderstood) book.
"End of History? Is that your new book or something? I haven't heard about it"
Great list.
I look forward to your extended reviews on the 1st 2 titles. I haven't read Di Resta yet, but Al-Gharbi was great. I was reading it during the US election in November, and everything seemed clearer to me.
Still, I think I agree that he is excessively woke, and I would like to see how you flesh that out. He reminds me of Veblen in that he takes the stance of a detached, objective socal scientist to deliver a half disguised semi religious jeremiad against the hypocrisy of his class.
On Marx, I recommend the highly readable book by Thomas Sowell, himself an ex-Marxist. Impressively, he presents Marx objectively throughout until unleashing on him in the final chapter. Actually, that's much the way I did it for my Marxism course at Oxford. My tutor liked me until that last class. It would be hard to find a more readable book about Marx than this one.
Intriguing selection, Dan Thanks for this and have a good 2025. All the best, John.
On your recommendation I guess I will have to read yet another book on Mars )
And 2 is good
Sorry mr heath
Pre ordered political epistemology I like his work
3 4 and 5 are indeed interesting
Haven't read the others
When's your book on cynicism coming out? 2025?
Surprised you've read a book on Marx in 2024. Is any more clarity required? His Communist Manifesto is pretty clear to me. While the vast majority of Western Academe is heavily influenced by him, many, though still critical of the West, now contradict him, particularly on Materialism versus Idealism. The scandal at your current university over Kathleen Stock epitomises this contradiction.
The Stock outrage is also why the upcoming book on Political Epistemology is of particular interest, so will you be reviewing it before its May 2025 publication?