Academia rightfully gets a lot of heat for this, but there are also egregious examples in the corporate world. I'm not talking about the DEI staff either. Elite overproduction means that the average Fortune 500 company has thousands of high IQ staff who do nothing but analytical masterbation and whatever other tasks exist solely because their company operates at a needlessly complex level. This is the equivalent of naval gazing in the academic world.
“In cases where the data can’t really tell us anything, I prefer to just welcome migrants due to the positive first order effects that they bring. We can figure out the political impacts later because we have no idea what they will eventually be.”
The economist Steve Landsburg, in his popular book _The Armchair Economist_, argues that one big reason that taxes are bad is that people don’t like to pay them. He doesn’t offer any data to explain the obvious. So, while it may be true that immigrants create positive first order effects, one way we know that mass non-white immigration is bad is that people don’t like alien phenotypes. One way we know this is true because pro-immigration advocates are constantly preaching about the alleged badness of rejecting alien phenotypes/racism. But because so many pro-immigration folks (even those who are “based”) have internalized the PC anti-racist ideology, they (including so-called economists) simply dismiss the massive externalities staring them in the face. In other words, mass non-white immigration doesn’t have to lead to “balkanization” to be bad (although the peaceful existence of racial enclaves is a kind of balkanization).
In addition, while often decrying Trump’s violation of democratic norms, pro-immigration folks ignore the preferences of the masses and endorse state-managed mass immigration that ignores the de jure ownership of the streets, roads, etc. by the natives. In this attitude, they are akin to progressive puritans banning and/or taxing cigarettes because they know better than the masses. Again, it is not “collectivism” or a violation of classical liberal tenets to defend de jure ownership. If you are a libertarian, the answer to “Who owns the streets?” can’t be the whole world—that would be communism. Therefore, respect for property rights means immigration by invitation only, not simply a border crossing. (And while decrying the “collectivism” of de jure ownership, some of these so-called libertarians have no trouble defending Israel and its bombing campaign against Gaza.)
The reality is that if we had had private ownership of streets, etc. all along, mass non-white immigration would have been impossible on the levels it exists today. Instead, the anti-racists prefer to violate de jure property to achieve their policy preferences—preferences primarily built upon an anti-racist ideology, not just economics (although economics that ignores externalities is also bad).
This articles calls to mind, entire books written with references to questionable data. Although the title of the book escapes me, my husband - in ultra liberal and data freak- read and my life has been hell since. The indicated that conservatives desire and seek answers, whereas liberals are content to say”I have no idea and because you’re assuming something just shows that my being comfortable with the unknown shows my political superiority “. Until of course it’s his turn to speculate on why this nation is so divided politically. It’s not a very generous, open-ended speculation I can tell you that. But he cites this book over and over,
Richard, only a few days ago you tweeted a single graph that supposedly "proved" that Europeans love immigration because, according to this single graph, they were relatively content with their "political system" (not further defined). You used this to dunk on conservatives critical of immigration who refused to "look at all the data" (again, you tweeted a single graph, with no source or context, as evidence for your own point).
I actually think this is a relatively easy thing to fix. Just reward researchers for tearing down claims by other academics. Indeed, that tends to be an important factor in distinguishing epistemically good and troubled fields.
IMO every major journal should reserve 2 slots for every paper they publish. The first slot goes to the best criticism submitted within 2 months of publication (so entirely methods/theoretical) which can be published along an author reply. The second slot (in empirical fields) goes to the first pre-registered study of similar power on the same subject. In fields relying on natural experiments one instead commits to publishing both the first similarly compelling analysis against and supporting the initial claim.
My youngest son is the academic. He was doing his M.S. in Psych. It's mostly research. He said there are very, very few topics allowed for research in the first place (for this course, at least). He got assigned a previous study on something he thought silly. His job was to expand on it. He looked through it and said the entire study was done incorrectly. So, basically he ended up having to do a paper on how to conduct the study properly in the first place, using the data given. This was his thesis paper. He was not happy, and says the same thing you say, which is that it's mostly b.s., a waste of time, and can be made to say whatever the author wants too often.
Thanks for writing this piece. I majored in Economics in college and took a number of advanced econometrics classes, including writing an econometric-based thesis. We were taught that this was proper methodology to use when researching a paper:
1. Hypothesize about a potential cause-effect to study
2. Gather data to measure this cause and effect
3. List out a series of controls for a regression model
4. Run the regression model
5. Write the paper on whatever results the regression model gives and compare to your hypotheses
Of course, if 4 doesn't return the results you want, well... didn't you forget to control for race? age? fixed effects? multicollinearity? There are so many plausible justifications for including / excluding controls or tweak the type of regression you're running that it's impossible to really pick the most appropriate / correct model. Because, of course, who's to say what is or isn't the "correct" model for the dataset.
I have friends who either don't think this happens so often or maybe that the solution is more rigorous oversight over these papers. Certainly, papers that went through an adversarial process where they had to share data, models, and had another researcher attempt to falsify their results, but failed to, would signal a robustness to the relationship they found. However, A) this would never happen (and the fact that it hasn't happened is an indictment of academia) and B) doesn't the fact that you can change results by slightly tuning a model signal that maybe these models aren't that useful?
The idea of regression models being able to tease out insights from large, complex datasets is alluring, but I think we have to be honest both that they are being misused and their ability to actually tell us that each year of schooling provides "$10,456" of income per year is bullshit.
Unfortunately even sibling birth order is a huge confounding variable. There's a huge benefit to being a first child and a super large pct of CEOs are first children.
Ah, Economics. You know, people with "Nobel prizes" go as far as saying things like "+4°C will only marginally affect the economy as only outdoors activities like agriculture, which represents a couple percent of GDP, will be impacted". Because as you know, agriculture as part of GDP perfectly represents its actual importance in our lives: economists don't eat I suppose :) A joke I like is "the brain is only 2% of the weight of an economist, therefore its ablation will only have a marginal impact on this economist's ability to function".
Focusing on a few important well-established truths is probably the right approach epistemically, but exactly the opposite of what writers tend to do to distinguish themselves or be interesting or seem intelligent.
I see conflicting studies about Charter schools. I suspect it is researchers confirming their priors.
Meanwhile, common sense tells me that giving parents choice is good, because they're motivated to make the right choice, and well-placed to evaluate their particular schools
Well written, if a tad depressing. The idea that academics should disclose their priors and not stray too far from the limits of the data is a good idea, but feels utopian given the incentive structure in academia.
I have also taught university students, but in computer science classes, and there I feel like my experience is quite different than yours. You really do get better at the skill of programming computers when some professor is assigning homework projects and making sure you actually follow through and do them. Sure you can teach yourself, but if you are very smart and only slightly motivated, the computer science college degree is a perfect fit.
Social science is not unique when it comes to the ‘fake data’ problem, by a long shot. I believe the first explosive article on the replication crisis occurred in the biomedical area. When a scientific field becomes important via money or political influence, the incentives for the scientists in that area will change - rendering the field less objective, less scientific. Some version of Goodhart’s Law might apply. For an outsider to that science I suppose relying heavily on “common sense” is better than just tallying up the outcomes of multiple studies. But what can be done to address the underlying problem of incentives?
Academia rightfully gets a lot of heat for this, but there are also egregious examples in the corporate world. I'm not talking about the DEI staff either. Elite overproduction means that the average Fortune 500 company has thousands of high IQ staff who do nothing but analytical masterbation and whatever other tasks exist solely because their company operates at a needlessly complex level. This is the equivalent of naval gazing in the academic world.
“In cases where the data can’t really tell us anything, I prefer to just welcome migrants due to the positive first order effects that they bring. We can figure out the political impacts later because we have no idea what they will eventually be.”
The economist Steve Landsburg, in his popular book _The Armchair Economist_, argues that one big reason that taxes are bad is that people don’t like to pay them. He doesn’t offer any data to explain the obvious. So, while it may be true that immigrants create positive first order effects, one way we know that mass non-white immigration is bad is that people don’t like alien phenotypes. One way we know this is true because pro-immigration advocates are constantly preaching about the alleged badness of rejecting alien phenotypes/racism. But because so many pro-immigration folks (even those who are “based”) have internalized the PC anti-racist ideology, they (including so-called economists) simply dismiss the massive externalities staring them in the face. In other words, mass non-white immigration doesn’t have to lead to “balkanization” to be bad (although the peaceful existence of racial enclaves is a kind of balkanization).
In addition, while often decrying Trump’s violation of democratic norms, pro-immigration folks ignore the preferences of the masses and endorse state-managed mass immigration that ignores the de jure ownership of the streets, roads, etc. by the natives. In this attitude, they are akin to progressive puritans banning and/or taxing cigarettes because they know better than the masses. Again, it is not “collectivism” or a violation of classical liberal tenets to defend de jure ownership. If you are a libertarian, the answer to “Who owns the streets?” can’t be the whole world—that would be communism. Therefore, respect for property rights means immigration by invitation only, not simply a border crossing. (And while decrying the “collectivism” of de jure ownership, some of these so-called libertarians have no trouble defending Israel and its bombing campaign against Gaza.)
The reality is that if we had had private ownership of streets, etc. all along, mass non-white immigration would have been impossible on the levels it exists today. Instead, the anti-racists prefer to violate de jure property to achieve their policy preferences—preferences primarily built upon an anti-racist ideology, not just economics (although economics that ignores externalities is also bad).
This articles calls to mind, entire books written with references to questionable data. Although the title of the book escapes me, my husband - in ultra liberal and data freak- read and my life has been hell since. The indicated that conservatives desire and seek answers, whereas liberals are content to say”I have no idea and because you’re assuming something just shows that my being comfortable with the unknown shows my political superiority “. Until of course it’s his turn to speculate on why this nation is so divided politically. It’s not a very generous, open-ended speculation I can tell you that. But he cites this book over and over,
Richard, only a few days ago you tweeted a single graph that supposedly "proved" that Europeans love immigration because, according to this single graph, they were relatively content with their "political system" (not further defined). You used this to dunk on conservatives critical of immigration who refused to "look at all the data" (again, you tweeted a single graph, with no source or context, as evidence for your own point).
I actually think this is a relatively easy thing to fix. Just reward researchers for tearing down claims by other academics. Indeed, that tends to be an important factor in distinguishing epistemically good and troubled fields.
IMO every major journal should reserve 2 slots for every paper they publish. The first slot goes to the best criticism submitted within 2 months of publication (so entirely methods/theoretical) which can be published along an author reply. The second slot (in empirical fields) goes to the first pre-registered study of similar power on the same subject. In fields relying on natural experiments one instead commits to publishing both the first similarly compelling analysis against and supporting the initial claim.
My youngest son is the academic. He was doing his M.S. in Psych. It's mostly research. He said there are very, very few topics allowed for research in the first place (for this course, at least). He got assigned a previous study on something he thought silly. His job was to expand on it. He looked through it and said the entire study was done incorrectly. So, basically he ended up having to do a paper on how to conduct the study properly in the first place, using the data given. This was his thesis paper. He was not happy, and says the same thing you say, which is that it's mostly b.s., a waste of time, and can be made to say whatever the author wants too often.
Thanks for writing this piece. I majored in Economics in college and took a number of advanced econometrics classes, including writing an econometric-based thesis. We were taught that this was proper methodology to use when researching a paper:
1. Hypothesize about a potential cause-effect to study
2. Gather data to measure this cause and effect
3. List out a series of controls for a regression model
4. Run the regression model
5. Write the paper on whatever results the regression model gives and compare to your hypotheses
Of course, if 4 doesn't return the results you want, well... didn't you forget to control for race? age? fixed effects? multicollinearity? There are so many plausible justifications for including / excluding controls or tweak the type of regression you're running that it's impossible to really pick the most appropriate / correct model. Because, of course, who's to say what is or isn't the "correct" model for the dataset.
I have friends who either don't think this happens so often or maybe that the solution is more rigorous oversight over these papers. Certainly, papers that went through an adversarial process where they had to share data, models, and had another researcher attempt to falsify their results, but failed to, would signal a robustness to the relationship they found. However, A) this would never happen (and the fact that it hasn't happened is an indictment of academia) and B) doesn't the fact that you can change results by slightly tuning a model signal that maybe these models aren't that useful?
The idea of regression models being able to tease out insights from large, complex datasets is alluring, but I think we have to be honest both that they are being misused and their ability to actually tell us that each year of schooling provides "$10,456" of income per year is bullshit.
Unfortunately even sibling birth order is a huge confounding variable. There's a huge benefit to being a first child and a super large pct of CEOs are first children.
Ah, Economics. You know, people with "Nobel prizes" go as far as saying things like "+4°C will only marginally affect the economy as only outdoors activities like agriculture, which represents a couple percent of GDP, will be impacted". Because as you know, agriculture as part of GDP perfectly represents its actual importance in our lives: economists don't eat I suppose :) A joke I like is "the brain is only 2% of the weight of an economist, therefore its ablation will only have a marginal impact on this economist's ability to function".
Good science still requires what c.s. peirce called abductive inference. Result/observation + rule > case. No amount of data can make up for this.
Focusing on a few important well-established truths is probably the right approach epistemically, but exactly the opposite of what writers tend to do to distinguish themselves or be interesting or seem intelligent.
I see conflicting studies about Charter schools. I suspect it is researchers confirming their priors.
Meanwhile, common sense tells me that giving parents choice is good, because they're motivated to make the right choice, and well-placed to evaluate their particular schools
Well written, if a tad depressing. The idea that academics should disclose their priors and not stray too far from the limits of the data is a good idea, but feels utopian given the incentive structure in academia.
I have also taught university students, but in computer science classes, and there I feel like my experience is quite different than yours. You really do get better at the skill of programming computers when some professor is assigning homework projects and making sure you actually follow through and do them. Sure you can teach yourself, but if you are very smart and only slightly motivated, the computer science college degree is a perfect fit.
Social science is not unique when it comes to the ‘fake data’ problem, by a long shot. I believe the first explosive article on the replication crisis occurred in the biomedical area. When a scientific field becomes important via money or political influence, the incentives for the scientists in that area will change - rendering the field less objective, less scientific. Some version of Goodhart’s Law might apply. For an outsider to that science I suppose relying heavily on “common sense” is better than just tallying up the outcomes of multiple studies. But what can be done to address the underlying problem of incentives?
GIGO