Comment by wyager | Hacker News Reader

Comment by wyager | Hacker News Reader

Parent Full thread

You’re making the same exact mistake that education activists have made for the last 60 years.

Education does not significantly improve outcomes. Period.

What people observe is that people who went to top-tier schools end up being very successful, and make the (totally false) inference that top-tier schools make you successful.

In reality, the people who were almost certainly going to be successful anyway get into the top-tier schools, partially on merit and partially because this is the legitimation process for the upper class in the US.

If you stuff a million poor people into the top-tier schools, this won’t actually fix anything, because A) those people are much less likely to succeed no matter what B) you can’t actually have a million people in the upper class.

We’re currently in the throes of dealing with B). Turchin calls this “elite overproduction”. You have millions of people who (think they) went through the process to become upper class (going to college), but it turns out that was just a signifier - a signifier that’s been totally destroyed by the push to send everyone to college.

> Education does not significantly improve outcomes. Period.

Define "significant".

People who don't graduate high school do much worse than people who graduate from college on average.

Even if you only graduate high school, you'll (on average) make half as much over your lifetime as someone who graduated with a Master's Degree [1].

Right now (on average) - people who dropped out of college make nearly 1/3rd as much as people who graduated with a Master's Degree [2].

You're also 36x more likely to be in prison [3]. Only ~10% of the >25 population has not graduated high school. Despite that, 80% of prisoners have not graduated from high school. ~18% of high school drop outs are in prison. Less than 0.5% of the rest of the population is in prison.

You're also almost 50% more likely to be obese (mostly because you're more likely to be poor and live somewhere un-walkable) [4].

You also have a ~13% shorter lifespan (9 less years) [5].

Sure - an education isn't a license to be in the top 1% or have a perfect life - but it certainly helps...

[1] https://www.aplu.org/projects-and-initiatives/college-costs-...

[2] https://study.com/articles/How_Much_More_Do_College_Graduate...

[3] https://slspotlight.com/opinion/2018/03/19/high-school-dropo...

[4] http://educationbythenumbers.org/content/education-might-be-...

[5] https://www.tc.columbia.edu/articles/2005/october/high-schoo...

Those data points don't disagree with wyager's point.

Both the the Signalling Theory they are advocating and the Human Capital Development (HCD) that I assume you are advocating would make the same predictions on those life indicators you listed.

The difference between the theories is which causes which: does consuming education cause people to have more of the traits that cause good life outcomes OR do the traits that cause good life outcomes also cause people to consume more education?

Pure HCD theory says it is only the first. Pure signalling says only the second.

The two can be mixed. What mix people believe in is important as the two theories result in very different policy goals for our education system.

But both agree that people with more education are going to earn more, commit fewer crimes, experience fewer crimes, be healthier, live longer, etc, etc, etc.

All mere correlations, which don't refute GP's claim.

When you make a bold claim that "Education does not significantly improve outcomes. Period." and correlations with every positive outcome are very high - the burden is on you to prove a non-causal link, not me.

But of course, proving causation is near impossible. So people love to make bold claims with no data and just say "prove me".

> and correlations with every positive outcome are very high

I explained why in the top-level post. People who were going to get positive outcomes anyway are good at getting into good schools.

> the burden is on you to prove a non-causal link, not me.

This isn’t Reddit - I wouldn’t drop some nonsense about made up “burden of proof” and expect anyone to buy it. It’s also rude to make objections like this when you clearly haven’t either considered the point you’re responding to (of course people who went to college do better in life - think about why that might be, under my model, and refute that if you can) or actually done the research into the presence or absence of a causal link.

There are 30 years of twin studies demonstrating conclusively that education is not an especially strong predictive factor in adult attainment.

I agree about the signifier, but still think education has a real value outside the pure signalling. You need experts and for that you need schools/colleges because in most jobs you will not learn wide enough nor deep enough (exceptions are possible).

Yes, you need experts, but the people who are going to really benefit from expert-level training were probably already going to go to college anyway. Doctors, lawyers, scientists, etc are part of the upper class in one way or another.

A huge chunk of "doctors, lawyers, scientists" come from the middle class or lower, though.

Yes, class mobility exists and is significant. A large fraction of high-intelligence people (which is a massive causal predictor of high class) come from parents of middling intelligence and class, just as a consequence of the ratio of generational intelligence drift to the width of the population intelligence distribution.

https://www.gov.uk/government/news/elitism-in-britain-2019

How are you defining education and significantly? How does education not provide you with the skills to be on the internet or language? How would you have faired if you were raised in a remote part of the world without structured or familial education?

I mean this in good spirit, is there proof that societal classes are bound by size?

Reply on news.ycombinator.com