Growing up in a neighborhood in which working adults are normal probably makes a big difference. What children see and experience as they grow up becomes “normal” to them. More often than not, they will do what is “normal” to them when they become adults.
It has always seemed to me down the years that - in mainstream media reporting of social science research - misplaced conflation of correlation with causation is the norm and not the exception. The resulting public consciousness ocean of what Twain called "the things you know for sure that ain't so" has been incalculably large.
There are a variety of clever tests to help show that a correlated relationship is likely causal in one direction and not due to confounding hidden variables, and I don't think an ordinary paper on a less politically charged topic could get far in the review process without doing those tests and showing some strong results. A good meta-test for whether a study is bunk is whether it gets a pass on that level of rigor. In the case of Chetty he has plausible deniability since his separated-snapshots approach is not amenable to that kind of more rigorous analysis, but what that should mean is that someone else gets to test his results against, say, a database of whatever standardized test scores were recorded for the studied individuals. There are now even more clever ways to do all that while protecting personal information from leaking to researchers without authorization, but the truth is everyone already suspects the the results will prove politically incorrect, not to mention embarrassing for Chetty and his fans, and so everyone is going to parrot whatever excuses necessary to assert that of course such a thing should never and could never be done because reasons.
Interesting comment on research methodologies but I was talking of media reports rather than the researches themselves. I haven't wasted my time with MSM 'news' for years now so I'm mostly going back quite a bit but for many many years I used to wake up to BBC "researchers have shown"-type stuff of the kind I meant.
My experience is that main stream 'reporters' who understand even the basics of statistics and statistical analysis are incredibly few and far between.
Tjht doesn't mean the study they 'report' on is good or bad, it just means they're talking about something they manifestly don't understand.
Whether his results are correct or not, can you explain why you are so certain his results regarding unemployment in the neighborhood will prove false?
As I'm sure you know, Sailer has looked at this guy's work but from the perspective of a (biological) realist, allowing him to come to some more parsimonious conclusions.
It's not just that though, and it also goes beyond adding jn selection effects. Interpreting some of Chetty's geographic results requires one to know some history and facts about local differences and actually understand that "ceteris is not paribus", that one often can't take places even within the same country as interchangeable economies with the only important causal distinction being local goverment policies. Sometimes a sector has a disproportionate significance in a local economy and so there are such things as local booms and busts unrelated to policy but that will coincidentally show up as huge signals when one takes Chetty's snapshots approach. The remarkable thing about Sailer is his breadth and memory for such cases and practical unique ability to combine an application of high levels of both erudition and common sense. I pray he lives long, there are presently no obvious worthy successors.
I don’t think this is just a selection effect. Anecdotally, the parents of your friends can have a massive impact on showing what kind of life is possible, especially when juxtaposed against your own parents’ lives if they are less well off.
The selection effect is the your parents chose to live next to the parents of your friends, so that you would grow up with better friend and better examples of their parents. Parents eager and able to make such choices are not random, so neither are their kids.
Arnold in this case I don't think selection is playing much of a role, he is looking at mostly unanticipated changes in community employment rates, and this is consistent with "China shock" deindustrialization effects that Autor and co-authors have identified. There are implications here for trade policy, which has given rise to what Rodrik calls the politics of anger. Some discussion of this here (though you may not like the proposed solutions):
"unanticipated changes in community employment rates"
What would be an anticipated change in employment rates- in either direction? Consider the hypothetical of a gentrifying neighborhood that was, 20 years prior, full of welfare recipients and criminals where the unemployment rate was 50% but over 20 years the previous residents died off and moved away as they were priced out by employed people who did the gentrifying. Would Chetty's study call that anticipated or unanticipated? Would it be tracking the residents of 20 years prior who moved away from the neighborhood or would it simply consider the neighborhood itself as the entity over the entire 20 years?
That was largely what I was coming to write as well.
I would agree with Mr Kling that Chetty isn't proving what he sets out to prove but that doesn't mean he isn't proving something that the Pro-Globalists and Free Trade Fanatics don't want to recognize.
A stable society needs opportunities for people of all cognitive and physical ability levels to make meaningful contributions to their communities. The dream of creating a 'clean' economy in the US that was primarily based on IP creators supported by a wide range of what amount to no more than 'personal assistants' is really the recipe for a society as unstable as the old feudal monarchies of Europe or any of the current petro-states. The high-low-no-middle mix means there are a wide range of people slotted into low-skill jobs for which they are largely overqualified and underpaid. This creates a larger than normal unemployable underclass even though they may be suitably qualified for low-skill work. The current Didn't Earn It 'fix' is to pound middle-skilled square pegs into high-qualification round holes, which only satisfies the cognitive elite with the belief they are 'doing something' about the social breakdown going on all around them.
Tax motivations aside, I find it remarkable that philanthropy is well-entrenched in America - and the list of ways people will find to fund causes related to low SES is mind-bogglingly long and varied - but the one thing that is really off the table, would get you laughed out of the room - is giving people jobs.
The problem with all of these "magic money wand" over educated idiots is they fail to recognize that *unless the people implementing the solutions are responsible for the intended outcome, they'll just give the money to their cronies or waste it.*
Accountability is the only thing we should focus on, because at present there isn't any.
I'm not sure I've heard Chetty's results for kindergarten teachers but there are plenty who have found similar short-term gains. Other studies have suggested these gains fade back to like their underperforming peers.
As for unemployment in the neighborhood, this seems rather obvious. It is what I would assume in the absence of contradictory evidence.
The other factor I would bet is as strong or stronger is growing up in a one vs two parent household.
Growing up in a neighborhood in which working adults are normal probably makes a big difference. What children see and experience as they grow up becomes “normal” to them. More often than not, they will do what is “normal” to them when they become adults.
It has always seemed to me down the years that - in mainstream media reporting of social science research - misplaced conflation of correlation with causation is the norm and not the exception. The resulting public consciousness ocean of what Twain called "the things you know for sure that ain't so" has been incalculably large.
There are a variety of clever tests to help show that a correlated relationship is likely causal in one direction and not due to confounding hidden variables, and I don't think an ordinary paper on a less politically charged topic could get far in the review process without doing those tests and showing some strong results. A good meta-test for whether a study is bunk is whether it gets a pass on that level of rigor. In the case of Chetty he has plausible deniability since his separated-snapshots approach is not amenable to that kind of more rigorous analysis, but what that should mean is that someone else gets to test his results against, say, a database of whatever standardized test scores were recorded for the studied individuals. There are now even more clever ways to do all that while protecting personal information from leaking to researchers without authorization, but the truth is everyone already suspects the the results will prove politically incorrect, not to mention embarrassing for Chetty and his fans, and so everyone is going to parrot whatever excuses necessary to assert that of course such a thing should never and could never be done because reasons.
Interesting comment on research methodologies but I was talking of media reports rather than the researches themselves. I haven't wasted my time with MSM 'news' for years now so I'm mostly going back quite a bit but for many many years I used to wake up to BBC "researchers have shown"-type stuff of the kind I meant.
My experience is that main stream 'reporters' who understand even the basics of statistics and statistical analysis are incredibly few and far between.
Tjht doesn't mean the study they 'report' on is good or bad, it just means they're talking about something they manifestly don't understand.
Whether his results are correct or not, can you explain why you are so certain his results regarding unemployment in the neighborhood will prove false?
"If the result of this research is that policy makers try to increase employment rates in poor areas, I see no harm"
The harm would be whatever money was wasted and whatever regulatory deadweight loss there was.
On anti-poverty programs which have net positive impacts on the poor and their families, the wise saying is, "The only thing that works, is work."
"Magic Dirt vs. Tragic Dirt" as the great Sailer puts it.
As I'm sure you know, Sailer has looked at this guy's work but from the perspective of a (biological) realist, allowing him to come to some more parsimonious conclusions.
It's not just that though, and it also goes beyond adding jn selection effects. Interpreting some of Chetty's geographic results requires one to know some history and facts about local differences and actually understand that "ceteris is not paribus", that one often can't take places even within the same country as interchangeable economies with the only important causal distinction being local goverment policies. Sometimes a sector has a disproportionate significance in a local economy and so there are such things as local booms and busts unrelated to policy but that will coincidentally show up as huge signals when one takes Chetty's snapshots approach. The remarkable thing about Sailer is his breadth and memory for such cases and practical unique ability to combine an application of high levels of both erudition and common sense. I pray he lives long, there are presently no obvious worthy successors.
I don’t think this is just a selection effect. Anecdotally, the parents of your friends can have a massive impact on showing what kind of life is possible, especially when juxtaposed against your own parents’ lives if they are less well off.
The selection effect is the your parents chose to live next to the parents of your friends, so that you would grow up with better friend and better examples of their parents. Parents eager and able to make such choices are not random, so neither are their kids.
Sort of like "Middle class families own homes. If we make it easier to own a home more people will become middle class." How did that work out?
Arnold in this case I don't think selection is playing much of a role, he is looking at mostly unanticipated changes in community employment rates, and this is consistent with "China shock" deindustrialization effects that Autor and co-authors have identified. There are implications here for trade policy, which has given rise to what Rodrik calls the politics of anger. Some discussion of this here (though you may not like the proposed solutions):
https://internationalbanker.com/finance/social-support-financial-architecture-proposal-reform/
"unanticipated changes in community employment rates"
What would be an anticipated change in employment rates- in either direction? Consider the hypothetical of a gentrifying neighborhood that was, 20 years prior, full of welfare recipients and criminals where the unemployment rate was 50% but over 20 years the previous residents died off and moved away as they were priced out by employed people who did the gentrifying. Would Chetty's study call that anticipated or unanticipated? Would it be tracking the residents of 20 years prior who moved away from the neighborhood or would it simply consider the neighborhood itself as the entity over the entire 20 years?
That was largely what I was coming to write as well.
I would agree with Mr Kling that Chetty isn't proving what he sets out to prove but that doesn't mean he isn't proving something that the Pro-Globalists and Free Trade Fanatics don't want to recognize.
A stable society needs opportunities for people of all cognitive and physical ability levels to make meaningful contributions to their communities. The dream of creating a 'clean' economy in the US that was primarily based on IP creators supported by a wide range of what amount to no more than 'personal assistants' is really the recipe for a society as unstable as the old feudal monarchies of Europe or any of the current petro-states. The high-low-no-middle mix means there are a wide range of people slotted into low-skill jobs for which they are largely overqualified and underpaid. This creates a larger than normal unemployable underclass even though they may be suitably qualified for low-skill work. The current Didn't Earn It 'fix' is to pound middle-skilled square pegs into high-qualification round holes, which only satisfies the cognitive elite with the belief they are 'doing something' about the social breakdown going on all around them.
Tax motivations aside, I find it remarkable that philanthropy is well-entrenched in America - and the list of ways people will find to fund causes related to low SES is mind-bogglingly long and varied - but the one thing that is really off the table, would get you laughed out of the room - is giving people jobs.
So, aren't we basically comparing a single working mom household with a mom and a working dad household?
Sailer: Snarking is not "delving. :)
I'll admit to a bias. A Radical Centrist [https://thomaslhutcheson.substack.com/] will take offense at the curl of lip when Sailer snarls, "centrist."
The problem with all of these "magic money wand" over educated idiots is they fail to recognize that *unless the people implementing the solutions are responsible for the intended outcome, they'll just give the money to their cronies or waste it.*
Accountability is the only thing we should focus on, because at present there isn't any.
I'm not sure I've heard Chetty's results for kindergarten teachers but there are plenty who have found similar short-term gains. Other studies have suggested these gains fade back to like their underperforming peers.
As for unemployment in the neighborhood, this seems rather obvious. It is what I would assume in the absence of contradictory evidence.
The other factor I would bet is as strong or stronger is growing up in a one vs two parent household.