Emil, I'm stumped by why you just seem to scan p-values to evaluate studies. Can you explain? How would you know anything about a study or its meaning from the p-values?
In my experience, the problem with most academic research is validity, and valid application. P-values don't matter if the research is invalid. For example, I just read a widely cited study of the benefits of cutting off girls' breasts on transgender grounds (girls 13 and up). The study is just giving a questionnaire for "chest dysphoria".
The questionnaire *assumes the girls still had breasts*. It asks if they "bind" at night, if other people misgender them because of their breasts, etc.
The girls trivially scored at the floor (Never) on the scale. Improvement.
Obviously that's invalid. Like in social psychology, we cited an example of a study of "Denial of Environmental Realities" (by major researchers at NYU). It's a questionnaire with items like "The Earth is like a crowded spaceship with limited room and resources" and vague prophecies about doom (not even specific enough to say climate). That's an analogy and a prophecy, not reality.
A working paper on "misinformation", already in the media, used arbitrarily selected websites (chosen by the authors) to judge Twitter users' linking behavior, instead of taking the websites users linked to and going from there. This means we have no idea which websites users linked to beyond the authors' arbitrary and rigged choices. We also don't know anything about misinformation here since the authors didn't identify any or rate any websites on that dimension. Instead they had random people answer one item of how much they trusted a domain, 5-point scale, and whether they were familiar with the domain. They kept the ratings when people said they didn't know the domain. All smaller sites were scored as "misinformation sites", and all big legacy leftist media were rated the opposite...
I don't know that these types of invalidity would map to those legal studies. Effect size is often small, and I think the assumption that you can take a finding from say 20 years ago, something like the behavior of judges (much less in one region or state), and apply it prospectively and get the same results is a hell of an assumption. Culture changes. Diversity hiring alone could nuke any behavioral pattern, just from cultural differences before we get to merit or intelligence difference.
The problem with citing any research is that no one reads the studies, at least not carefully. They don't read them with exogenous intelligence, and seem to assume that a study can't be fundamentally invalid. The elites don't read the ivermectin studies, call for retraction of ivermectin studies just because others were retracted (in at least one case a study was retracted just because it was an ivermectin study, which creates interesting snowball dynamics). They don't pay attention to why some people weren't interested in the vaccines – unknown long-term safety of a brand new pharmaceutical in a weak cost-benefit context. Instead they'll spend years doing studies asking about microchips, 5G, and modification of DNA.
They don't pay attention to details, have no rigor when something is politicized. Since the elites in our culture are political, that pretty much ruins the idea of trusting them (we could have a political system where people are much less "political" in the sense I mean, but we have this insane culture).
Philosopher Jason Brennan is my first exposure to this idea. He wrote the book "Against Democracy" a few years ago. He's a libertarian, so not the usual angle. Still, he overestimates the accuracy of social science and economics "findings" and consensuses. But the last two years kill the idea – politics is too corrosive, for leftists especially. They're telling us Fetterman's mind works, he'll heal, women can compete with men in sports, and that Republican rhetoric whispers to crazy people like Sauron through the ring.
I favour a third seat of legislative power sitting above all others - a "citizens house" where members (suitably qualified - over 30, no criminal record, heterosexual, married, natural born citizens of 3+ generations standing etc) are elected by lot (sortition) for set terms, and required to conduct their own review of all legislation being passed, and senior appointments etc.. with the power to knock back any that they are not confident in, and even dismiss the government of the day if required.
It would be much more immune to corruption, and able to keep all other levels of law and politics straight. It would be a body severely unlikely to have ever passed our current immigration laws, or allowed them to stand.
On the question of whether more or less democracy is required, the answer is certainly, that movement in EITHER direction is preferable to what we have endured the last 100 years until today. Either markedly more democracy as configured like the above, where the populace get their say on every piece of legislation, or markedly less, where a good dictator of the folk could cut out the rot, without the hinderance of parliament and elite serving, anti-folk media, and the idiots programmed to support them.
Great article. I agree that democracy is deeply flawed, a fact that the Ancient Greeks have already known. I more and more believe that an actual aristocracy that leaves the commoners alone in most aspects, collecting only a few taxes, would be preferable to both, a democracy run by those promising short-sighted solutions to get reelected, and a technocracy which tries to social engineer us toward some utopia (that we would probably consider dystopian).
Demarchy is a governance structure that deserves more attention and experimentation. The 10% less democracy should be administered through a demarchic model rather than through politically active elites.
Emil, I'm stumped by why you just seem to scan p-values to evaluate studies. Can you explain? How would you know anything about a study or its meaning from the p-values?
In my experience, the problem with most academic research is validity, and valid application. P-values don't matter if the research is invalid. For example, I just read a widely cited study of the benefits of cutting off girls' breasts on transgender grounds (girls 13 and up). The study is just giving a questionnaire for "chest dysphoria".
The questionnaire *assumes the girls still had breasts*. It asks if they "bind" at night, if other people misgender them because of their breasts, etc.
The girls trivially scored at the floor (Never) on the scale. Improvement.
Obviously that's invalid. Like in social psychology, we cited an example of a study of "Denial of Environmental Realities" (by major researchers at NYU). It's a questionnaire with items like "The Earth is like a crowded spaceship with limited room and resources" and vague prophecies about doom (not even specific enough to say climate). That's an analogy and a prophecy, not reality.
A working paper on "misinformation", already in the media, used arbitrarily selected websites (chosen by the authors) to judge Twitter users' linking behavior, instead of taking the websites users linked to and going from there. This means we have no idea which websites users linked to beyond the authors' arbitrary and rigged choices. We also don't know anything about misinformation here since the authors didn't identify any or rate any websites on that dimension. Instead they had random people answer one item of how much they trusted a domain, 5-point scale, and whether they were familiar with the domain. They kept the ratings when people said they didn't know the domain. All smaller sites were scored as "misinformation sites", and all big legacy leftist media were rated the opposite...
I don't know that these types of invalidity would map to those legal studies. Effect size is often small, and I think the assumption that you can take a finding from say 20 years ago, something like the behavior of judges (much less in one region or state), and apply it prospectively and get the same results is a hell of an assumption. Culture changes. Diversity hiring alone could nuke any behavioral pattern, just from cultural differences before we get to merit or intelligence difference.
The problem with citing any research is that no one reads the studies, at least not carefully. They don't read them with exogenous intelligence, and seem to assume that a study can't be fundamentally invalid. The elites don't read the ivermectin studies, call for retraction of ivermectin studies just because others were retracted (in at least one case a study was retracted just because it was an ivermectin study, which creates interesting snowball dynamics). They don't pay attention to why some people weren't interested in the vaccines – unknown long-term safety of a brand new pharmaceutical in a weak cost-benefit context. Instead they'll spend years doing studies asking about microchips, 5G, and modification of DNA.
They don't pay attention to details, have no rigor when something is politicized. Since the elites in our culture are political, that pretty much ruins the idea of trusting them (we could have a political system where people are much less "political" in the sense I mean, but we have this insane culture).
Philosopher Jason Brennan is my first exposure to this idea. He wrote the book "Against Democracy" a few years ago. He's a libertarian, so not the usual angle. Still, he overestimates the accuracy of social science and economics "findings" and consensuses. But the last two years kill the idea – politics is too corrosive, for leftists especially. They're telling us Fetterman's mind works, he'll heal, women can compete with men in sports, and that Republican rhetoric whispers to crazy people like Sauron through the ring.
I favour a third seat of legislative power sitting above all others - a "citizens house" where members (suitably qualified - over 30, no criminal record, heterosexual, married, natural born citizens of 3+ generations standing etc) are elected by lot (sortition) for set terms, and required to conduct their own review of all legislation being passed, and senior appointments etc.. with the power to knock back any that they are not confident in, and even dismiss the government of the day if required.
It would be much more immune to corruption, and able to keep all other levels of law and politics straight. It would be a body severely unlikely to have ever passed our current immigration laws, or allowed them to stand.
On the question of whether more or less democracy is required, the answer is certainly, that movement in EITHER direction is preferable to what we have endured the last 100 years until today. Either markedly more democracy as configured like the above, where the populace get their say on every piece of legislation, or markedly less, where a good dictator of the folk could cut out the rot, without the hinderance of parliament and elite serving, anti-folk media, and the idiots programmed to support them.
Great article. I agree that democracy is deeply flawed, a fact that the Ancient Greeks have already known. I more and more believe that an actual aristocracy that leaves the commoners alone in most aspects, collecting only a few taxes, would be preferable to both, a democracy run by those promising short-sighted solutions to get reelected, and a technocracy which tries to social engineer us toward some utopia (that we would probably consider dystopian).
Demarchy is a governance structure that deserves more attention and experimentation. The 10% less democracy should be administered through a demarchic model rather than through politically active elites.