What comes to kind immediately is the tendency of leftists (and many sociologists) to criticize the quantitative nature of Econ in the opposite way. They often claim that Econ uses fancy equations and tables to give an intellectual veneer to their naked promotion of the interests of capital.
The next time any leftist or sociologist says that to you, remind them how Hayek and Mises were not only "far right" (by their standards) but also strident critics of high levels of mathematicization.
I share some of the views but I think the overall picture of the discipline is conveys is not representative. I wrote an answer. I hope some people find it interesting: https://substack.com/@lsage1/p-212004938
That’s a very good answer. I think we mostly agree and simply have different perspectives: me as an outsider looking at the broad picture emerging from the field, and you as someone involved in the minority part of the field doing the kind of work I think should be done more often.
Thanks for the kind words and for reposting! Yes indeed. It’s just sad for me to realize that from the outsider perspective, this kind of sociology is invisible…
I did not mean to say that it is invisible, but that it is not the centre of gravity of the discipline. Having said that, our perspectives might give us different impressions of where that centre of gravity lies. I totally agree with you that I might not appreciate the internal recognition of quantitative sociology as well as an insider would, but at the same time, an insider in that subfield could also potentially overestimate its importance. In any case, I found your piece an interesting counterpoint and, if anything, I hope you are right and that quantitative sociology already has more internal standing than I was suggesting.
Interesting read. Having some knowledge in both areas I’d argue that it really is difficult to divorce political paradigms from inquiry. It’s all about how different group behaviours would be seen. While Marx argues that conflict is necessary and questions those with power, others such as Durkheim take a more conservative approach justifying what already exists. This is political.
Economics started with a very fundamental belief in demand, supply and the unshakable belief in markets. It simplified the world, and the behaviours within it too much. Over the years these beliefs have been questioned through all sorts of economic paradigm shifts. Keynes and Marx amongst others. While Marx was not really accepted by economists really, Keynes was. Under specific circumstances Keynes was right, but the 1970s showed some crucial limitations. Game theory and behavioural economics also questioned the basics of traditional economics. One may look at these changes as being political, and they probably are. The personal is political.
What I’ve always liked about the social sciences is the capacity to understand that there is nuance. Seldom, where humans are involved, can we find a definitive answer or explanation for our behaviours. They are only right to an extent. The economic modelling which aims to quantify, with more assumptions than you can poke a stick at, is more wrong than most realise. It is likely that any quantitative sociology results are more correct, but methodology can be questioned too. I’d argue that both have their place though. While we aren’t rational a lot of the time, economics is still relevant. Different sociological explanations likewise.
Nuance? Strange term. The human "sciences" - sociology, economics, anthropology, historiography, psychology, psychoanalysis - have been a disaster. Why? Because policy makers LISTEN TO AND ACT UPON their research findings.
But they produce no scientifically valid data, BECAUSE WHAT THEY DO IS NOT SCIENTIFIC. Causality proper has no purchase on human affairs either of individuals or collectives.
Causality belongs in Classical Newtonian mechanics AND ONLY THERE. The Bordieu Road crash is unsurprising. Sociology posing as science is great training for teaching because it forces the teacher to defer to other disciplines to be able to make sense of sociological material to his students.
Precisely this. This protean vacuity is precisely why I, as a historian, have resisted including history in the social sciences (historiography is not a separate discipline; it is the study of how historians do history which can include all sorts of theoretical frameworks but is never prescriptive in the sense you mean). We are human, and so studying ourselves can never be fully objective, even when compared to natural sciences that are less focused on experimental methodologies like geology. Humans (at least breathing ones) tend to mess with the results and are studying you while you’re studying them. I understand that Game Theory is designed to work around thst issue, but I too, am not fully convinced that it is a foolproof way of doing so. Frankly, I would not only be skeptical of a person who presents me with a formula for precisely predicting human behavior: I would assume that person is dangerous and an enemy of the human race. Having said that, I still think there is profit in studying the patterns of human behavior, but the “lessons” there are perforce broad and open to constant reinterpretation.
I formula is interpreted as a good predictor beyond its context and assumptions. That’s humans too. Humans aren’t perfect in so many ways. We just have to look at these explanations as imperfect but not useless.
I would argue that we think we know history, but there are always imperfections in our explanations there. Physical evidence certainly helps, but the full meaning and understanding will, in part, be incomplete. We keep finding more. That’s a good thing.
Don, I’d say you look at this as though studying human behaviours is a waste of time because we can’t provide the laboratory conditions. I disagree somewhat. Being able to explain human behaviour is useful. The idea of social science is to bring what it can from science to study human behaviours. By no means am I saying it is perfect. Where I do think social scientists can fail is when they investigate an issue with a closed or predetermined mind. If that occurs then methodology is determined by the desired outcome. That is poor process which I don’t support.
I’d also like to add to the discussion that there is a bias to quantitative data these days. Data is not as good as many think it is. Even scientific data can be questioned. It should be if we are serious. This might mean we start to really engage with what is in front of us. Often what we look at may be a partial explanation (with the need for assumptions or data constraints). That is what I’m referring to in nuance.
Where we agree is that there have been some research findings that have lacked adequate peer review and other supporting research.
"it is interesting that this political shift also happened in economics, despite the discipline’s traditional association with capitalism and markets"
The discussion in Manzi's paper you link addresses this point specifically. The shift in the period is linked to social liberalism, not economic left thought. The economic slant didn't change much at all.
It should not surprise that economics departments in this period have become more socially liberal, because it is a key component of the Western liberal capitalist model.
I think anthropology is somewhat of a sister discipline. Psychology stands a bit apart in the holistic vs reductionist divide because its object of study is, in a sense, the individual, and the discipline as a whole does not have the same mission of connecting individual behaviour to social phenomena as sociology and economics. Psychology also has many subfields with very different traditions, some closely connected to the hard sciences, such as neuroscience, and others more willing to accept “social influences” in a way that resembles the sociological mindset, such as social psychology.
Thank God somebody write something making the distinctive nature of sociology plain.
I am so sick of people screaming about how universities have become infected by woke propaganda corrupted by critical theory. This is complete and yet blocks but it is a fashionable claim.
As I have repeatedly pointed out this is not only completely true of all science and engineering it is also true of economics and political science. The claim is so overblown...
Thanks for putting a stake in it's heart. Now I can just link to this article.
The intellectual legacy of the Frankfurt School—and the progressive, postmodern neo-Marxism that followed in its wake—has become one of the most corrosive forces in the Western world. It has infiltrated our institutions, weaponised grievance, degraded standards, distorted history and taught generations to regard their own civilisation with contempt. It must now be confronted without apology and stripped of its institutional power through democratic action, intellectual courage and the uncompromising restoration of truth, merit, responsibility and national confidence. Those who have abused public office or institutional authority to impose this ideology must face rigorous investigation and the fullest accountability permitted by law. There must be no violence and no persecution—but neither can there be silence, evasion or immunity from democratic judgment. The West has surrendered ground for long enough. It is time to take it back. Playing with the lives of millions is a very dangerous gamble.
There is a real fear that the failing epistemological standards of sociology will be pushed into other departments. The current vessel for this is the decolonise movement but if that fails then the sociologists with their humanities allies will just create a new vessel
I don't think we should be complacent, there is no certainty that academic science, mathematics and engineering can withstand this. By their nature the defenders are less political yet this is a political conflict.
Respectfully I think this is unduly paranoid. To take what I hope is the most obvious example - the chance that mathematics will be 'infected' by critical theory is less than the square root of fuck all )
It's not a matter of complacency. It's common sense. Critical theory always had an immediate and intimate relationship to sociology. There was always an agency to any subject with a string cultural dimension eg art criticism, literature criticism. Film studies.
But this factor just isn't there with science and engineering. Can a few people strain to create a connection. Absolutely. But it is impressive and candidly the overwhelming majority of scientists and engineers didn't give a shit.
Pretty sure you can sleep easy
To be clear. None of the above means that such disciplined are completely isolated from social concerns. There were scientist fellow travelers aplenty. But critical theory riding epistemic standards? Nah. You can sleep easy
Education is a mystery to me. This is a field where rigorous experiments could be run and a solid practice established, yet is constantly swept by unsubstantiated fads, which are then propagated to schools and cause no little damage (see "whole word" reading instruction).
If I were to dare a criticism of this excellent article it would be to question the direction of the casual relationship between the political shifts in sociology and left wing politics
Some left wing ideas entered sociology but they incubated and changed there. There is then an increasing exchange of ideas between left wing spaces and the sociological fields.
But I see sociology as more of the driver of this. The alignment of sociology with left thought in 2026 appears to me more because sociology has changed left wing thought. The constant feed of activist graduates into the political machinery would be where I would start research into this causal chain if I were a researcher. Left wing opinions align with current sociology because the left wing activist class is predominantly educated/initiated into those worldviews in academic departments that developed those ideas and the language to express them
Good essay. I was with you until you suggested economists can use statistical techniques to get “as good as experiment” causal data. But that’s just not true. That’s been tried for decades from SEM to DiD and these approaches have massive false positive problems, too many researchers degrees of freedom, and ignorance of near zero effect sizes preferring naive interpretation of p values. This isn’t a defense of sociology but Ive seen economists in my own area make a lot of faulty inferences based in such statistics. There’s no magic wand to convert correlation to causation. OLS works “fine” so long, as you say, you don’t assume causality. The newer statistics are often trendy toys with lots of problems but don’t produce more reliable causal inferences.
There definitely are subgroups within sociology that are doing rigorous work. I would call Peter Turchin's work very rigorous and empirically constrained, and sociological in nature. But it's clearly an outlier, and in Turchin's case he started in the hard sciences. But why that didn't become the dominant framework in sociology is an open question.
Very interesting, as always. I feel that in our own field, macroeconomics shares some of these problems in that we can't run experiments and we observe macro variables that are the result of complex interactions of millions of micro decisions. So far, attempts at models with predictive power have not been overly successful, perhaps because they have to be so simplified to have any tractability.
I've also experienced some social scientists who are more likely to not publish a finding if it disagrees with their politics which really feels counterproductive. It's good to keep your values, not your conclusions in the face of the data.
I would also like to state, for the record, that some of my best friends are sociologists (and political scientists).
I'm not quite sure what to make of the fact that right after noting how the reproducibility crisis is kind of a big deal since it suggests that even using the same data and the same methods can lead to wildly different results you then cite a study premised around an AI model. If the systems we're using are intrinsically flawed, technology based on these systems that reorganizes it in more arcane ways that are even harder to verify isn't the solution.
The distinction isn't really mathematical (all the Austrian School economists were skeptical of mathematical modelling). The real distinction is methodological individualism (the bottom up approach) vs methodological collectivism (the top down approach).
If you aren't starting your social science with methodological individualism, you're engaged in quackery. Social conventions/norms/habits/routines/etc are absolutely real things but can be described quite well with a methodologically individualist approach (begin with the idea of individuals economizing on social status, and treat social norms as things which compliance with/defection from accrues benefits/costs respectively).
Just as Public Choice Theory brought economists into the realm of Political Science, we need an economic incursion into sociology. An absolute REVERSAL of Amitai Etzioni's "Socio-Economics" - we don't need to make econ more like sociology, but rather sociology more like econ.
Right-wingers would do well to read more from Bourdieu.
Just like Marx, Bourdieu often feels like he is on to something, but at the same time takes the wrong approach.
I often think about these socialists and all the ways their ideas are different from the real world. (Yes, for real, I walk around thinking about these things daily while laughing to myself about what morons they were, and how so many academics still believe in them)
To Bourdieu, everything was about power and dominance. School, for example, is symbolic violence enacted by those with power.
But to most of us (who are living in the real world and not in sociological institutions) schools are how we teach young people what they need to know in society, in the way that best benefits both them and society.
Bourdieu actually wrote some of the funniest books I've read, but he probably didn't intend it that way. I still laugh every time I think about his distaste for how the working class had such low taste that they enjoyed 🤣Petula Clarke's song "Downtown". 😂
A great read. Thanks.
What comes to kind immediately is the tendency of leftists (and many sociologists) to criticize the quantitative nature of Econ in the opposite way. They often claim that Econ uses fancy equations and tables to give an intellectual veneer to their naked promotion of the interests of capital.
I did my PhD in game theory so... Yeah.
The next time any leftist or sociologist says that to you, remind them how Hayek and Mises were not only "far right" (by their standards) but also strident critics of high levels of mathematicization.
I share some of the views but I think the overall picture of the discipline is conveys is not representative. I wrote an answer. I hope some people find it interesting: https://substack.com/@lsage1/p-212004938
That’s a very good answer. I think we mostly agree and simply have different perspectives: me as an outsider looking at the broad picture emerging from the field, and you as someone involved in the minority part of the field doing the kind of work I think should be done more often.
Thanks for the kind words and for reposting! Yes indeed. It’s just sad for me to realize that from the outsider perspective, this kind of sociology is invisible…
I did not mean to say that it is invisible, but that it is not the centre of gravity of the discipline. Having said that, our perspectives might give us different impressions of where that centre of gravity lies. I totally agree with you that I might not appreciate the internal recognition of quantitative sociology as well as an insider would, but at the same time, an insider in that subfield could also potentially overestimate its importance. In any case, I found your piece an interesting counterpoint and, if anything, I hope you are right and that quantitative sociology already has more internal standing than I was suggesting.
Interesting read. Having some knowledge in both areas I’d argue that it really is difficult to divorce political paradigms from inquiry. It’s all about how different group behaviours would be seen. While Marx argues that conflict is necessary and questions those with power, others such as Durkheim take a more conservative approach justifying what already exists. This is political.
Economics started with a very fundamental belief in demand, supply and the unshakable belief in markets. It simplified the world, and the behaviours within it too much. Over the years these beliefs have been questioned through all sorts of economic paradigm shifts. Keynes and Marx amongst others. While Marx was not really accepted by economists really, Keynes was. Under specific circumstances Keynes was right, but the 1970s showed some crucial limitations. Game theory and behavioural economics also questioned the basics of traditional economics. One may look at these changes as being political, and they probably are. The personal is political.
What I’ve always liked about the social sciences is the capacity to understand that there is nuance. Seldom, where humans are involved, can we find a definitive answer or explanation for our behaviours. They are only right to an extent. The economic modelling which aims to quantify, with more assumptions than you can poke a stick at, is more wrong than most realise. It is likely that any quantitative sociology results are more correct, but methodology can be questioned too. I’d argue that both have their place though. While we aren’t rational a lot of the time, economics is still relevant. Different sociological explanations likewise.
Nuance? Strange term. The human "sciences" - sociology, economics, anthropology, historiography, psychology, psychoanalysis - have been a disaster. Why? Because policy makers LISTEN TO AND ACT UPON their research findings.
But they produce no scientifically valid data, BECAUSE WHAT THEY DO IS NOT SCIENTIFIC. Causality proper has no purchase on human affairs either of individuals or collectives.
Causality belongs in Classical Newtonian mechanics AND ONLY THERE. The Bordieu Road crash is unsurprising. Sociology posing as science is great training for teaching because it forces the teacher to defer to other disciplines to be able to make sense of sociological material to his students.
Precisely this. This protean vacuity is precisely why I, as a historian, have resisted including history in the social sciences (historiography is not a separate discipline; it is the study of how historians do history which can include all sorts of theoretical frameworks but is never prescriptive in the sense you mean). We are human, and so studying ourselves can never be fully objective, even when compared to natural sciences that are less focused on experimental methodologies like geology. Humans (at least breathing ones) tend to mess with the results and are studying you while you’re studying them. I understand that Game Theory is designed to work around thst issue, but I too, am not fully convinced that it is a foolproof way of doing so. Frankly, I would not only be skeptical of a person who presents me with a formula for precisely predicting human behavior: I would assume that person is dangerous and an enemy of the human race. Having said that, I still think there is profit in studying the patterns of human behavior, but the “lessons” there are perforce broad and open to constant reinterpretation.
I formula is interpreted as a good predictor beyond its context and assumptions. That’s humans too. Humans aren’t perfect in so many ways. We just have to look at these explanations as imperfect but not useless.
I would argue that we think we know history, but there are always imperfections in our explanations there. Physical evidence certainly helps, but the full meaning and understanding will, in part, be incomplete. We keep finding more. That’s a good thing.
Don, I’d say you look at this as though studying human behaviours is a waste of time because we can’t provide the laboratory conditions. I disagree somewhat. Being able to explain human behaviour is useful. The idea of social science is to bring what it can from science to study human behaviours. By no means am I saying it is perfect. Where I do think social scientists can fail is when they investigate an issue with a closed or predetermined mind. If that occurs then methodology is determined by the desired outcome. That is poor process which I don’t support.
I’d also like to add to the discussion that there is a bias to quantitative data these days. Data is not as good as many think it is. Even scientific data can be questioned. It should be if we are serious. This might mean we start to really engage with what is in front of us. Often what we look at may be a partial explanation (with the need for assumptions or data constraints). That is what I’m referring to in nuance.
Where we agree is that there have been some research findings that have lacked adequate peer review and other supporting research.
"it is interesting that this political shift also happened in economics, despite the discipline’s traditional association with capitalism and markets"
The discussion in Manzi's paper you link addresses this point specifically. The shift in the period is linked to social liberalism, not economic left thought. The economic slant didn't change much at all.
It should not surprise that economics departments in this period have become more socially liberal, because it is a key component of the Western liberal capitalist model.
Very interesting article.
If there is a continuum between economics and sociology, where do you see the other social sciences?
I think anthropology is somewhat of a sister discipline. Psychology stands a bit apart in the holistic vs reductionist divide because its object of study is, in a sense, the individual, and the discipline as a whole does not have the same mission of connecting individual behaviour to social phenomena as sociology and economics. Psychology also has many subfields with very different traditions, some closely connected to the hard sciences, such as neuroscience, and others more willing to accept “social influences” in a way that resembles the sociological mindset, such as social psychology.
Thank God somebody write something making the distinctive nature of sociology plain.
I am so sick of people screaming about how universities have become infected by woke propaganda corrupted by critical theory. This is complete and yet blocks but it is a fashionable claim.
As I have repeatedly pointed out this is not only completely true of all science and engineering it is also true of economics and political science. The claim is so overblown...
Thanks for putting a stake in it's heart. Now I can just link to this article.
The intellectual legacy of the Frankfurt School—and the progressive, postmodern neo-Marxism that followed in its wake—has become one of the most corrosive forces in the Western world. It has infiltrated our institutions, weaponised grievance, degraded standards, distorted history and taught generations to regard their own civilisation with contempt. It must now be confronted without apology and stripped of its institutional power through democratic action, intellectual courage and the uncompromising restoration of truth, merit, responsibility and national confidence. Those who have abused public office or institutional authority to impose this ideology must face rigorous investigation and the fullest accountability permitted by law. There must be no violence and no persecution—but neither can there be silence, evasion or immunity from democratic judgment. The West has surrendered ground for long enough. It is time to take it back. Playing with the lives of millions is a very dangerous gamble.
The claim is also somewhat a fear
There is a real fear that the failing epistemological standards of sociology will be pushed into other departments. The current vessel for this is the decolonise movement but if that fails then the sociologists with their humanities allies will just create a new vessel
I don't think we should be complacent, there is no certainty that academic science, mathematics and engineering can withstand this. By their nature the defenders are less political yet this is a political conflict.
Respectfully I think this is unduly paranoid. To take what I hope is the most obvious example - the chance that mathematics will be 'infected' by critical theory is less than the square root of fuck all )
It's not a matter of complacency. It's common sense. Critical theory always had an immediate and intimate relationship to sociology. There was always an agency to any subject with a string cultural dimension eg art criticism, literature criticism. Film studies.
But this factor just isn't there with science and engineering. Can a few people strain to create a connection. Absolutely. But it is impressive and candidly the overwhelming majority of scientists and engineers didn't give a shit.
Pretty sure you can sleep easy
To be clear. None of the above means that such disciplined are completely isolated from social concerns. There were scientist fellow travelers aplenty. But critical theory riding epistemic standards? Nah. You can sleep easy
Completely untrue
Fat fingers
Education is a mystery to me. This is a field where rigorous experiments could be run and a solid practice established, yet is constantly swept by unsubstantiated fads, which are then propagated to schools and cause no little damage (see "whole word" reading instruction).
If I were to dare a criticism of this excellent article it would be to question the direction of the casual relationship between the political shifts in sociology and left wing politics
Some left wing ideas entered sociology but they incubated and changed there. There is then an increasing exchange of ideas between left wing spaces and the sociological fields.
But I see sociology as more of the driver of this. The alignment of sociology with left thought in 2026 appears to me more because sociology has changed left wing thought. The constant feed of activist graduates into the political machinery would be where I would start research into this causal chain if I were a researcher. Left wing opinions align with current sociology because the left wing activist class is predominantly educated/initiated into those worldviews in academic departments that developed those ideas and the language to express them
Good essay. I was with you until you suggested economists can use statistical techniques to get “as good as experiment” causal data. But that’s just not true. That’s been tried for decades from SEM to DiD and these approaches have massive false positive problems, too many researchers degrees of freedom, and ignorance of near zero effect sizes preferring naive interpretation of p values. This isn’t a defense of sociology but Ive seen economists in my own area make a lot of faulty inferences based in such statistics. There’s no magic wand to convert correlation to causation. OLS works “fine” so long, as you say, you don’t assume causality. The newer statistics are often trendy toys with lots of problems but don’t produce more reliable causal inferences.
There definitely are subgroups within sociology that are doing rigorous work. I would call Peter Turchin's work very rigorous and empirically constrained, and sociological in nature. But it's clearly an outlier, and in Turchin's case he started in the hard sciences. But why that didn't become the dominant framework in sociology is an open question.
Very interesting, as always. I feel that in our own field, macroeconomics shares some of these problems in that we can't run experiments and we observe macro variables that are the result of complex interactions of millions of micro decisions. So far, attempts at models with predictive power have not been overly successful, perhaps because they have to be so simplified to have any tractability.
I've also experienced some social scientists who are more likely to not publish a finding if it disagrees with their politics which really feels counterproductive. It's good to keep your values, not your conclusions in the face of the data.
I would also like to state, for the record, that some of my best friends are sociologists (and political scientists).
I'm not quite sure what to make of the fact that right after noting how the reproducibility crisis is kind of a big deal since it suggests that even using the same data and the same methods can lead to wildly different results you then cite a study premised around an AI model. If the systems we're using are intrinsically flawed, technology based on these systems that reorganizes it in more arcane ways that are even harder to verify isn't the solution.
The distinction isn't really mathematical (all the Austrian School economists were skeptical of mathematical modelling). The real distinction is methodological individualism (the bottom up approach) vs methodological collectivism (the top down approach).
If you aren't starting your social science with methodological individualism, you're engaged in quackery. Social conventions/norms/habits/routines/etc are absolutely real things but can be described quite well with a methodologically individualist approach (begin with the idea of individuals economizing on social status, and treat social norms as things which compliance with/defection from accrues benefits/costs respectively).
Just as Public Choice Theory brought economists into the realm of Political Science, we need an economic incursion into sociology. An absolute REVERSAL of Amitai Etzioni's "Socio-Economics" - we don't need to make econ more like sociology, but rather sociology more like econ.
All people lust for power. The sociologists did it with claims of superior insight. They promise to be good masters, but they mean to be masters.
Right-wingers would do well to read more from Bourdieu.
Just like Marx, Bourdieu often feels like he is on to something, but at the same time takes the wrong approach.
I often think about these socialists and all the ways their ideas are different from the real world. (Yes, for real, I walk around thinking about these things daily while laughing to myself about what morons they were, and how so many academics still believe in them)
To Bourdieu, everything was about power and dominance. School, for example, is symbolic violence enacted by those with power.
But to most of us (who are living in the real world and not in sociological institutions) schools are how we teach young people what they need to know in society, in the way that best benefits both them and society.
Bourdieu actually wrote some of the funniest books I've read, but he probably didn't intend it that way. I still laugh every time I think about his distaste for how the working class had such low taste that they enjoyed 🤣Petula Clarke's song "Downtown". 😂
Always has been?