[wojt_eu]

This shows correlation but not causation. Hypothetically: people with certain background (social class) and traits (IQ) are more likely to go to college, and they would fare as well or better without college. I'd be very interested in stats of lifetime earnings difference that was corrected for selection bias.

[uberman]

Would a 1% sample of the US population that included level of education and earnings do the trick for you?

https://factfinder.census.gov/faces/tableservices/jsf/pages/...

It is not going to have variables like "Social status of your parents" or "IQ" but it does have a rather extensive list of demographic variables to control for.

Let me know what you uncover.

[jessriedel]

IQ, conscientiousness, and parental social status are probably the three biggest determiners of success, presumably none of which are being controlled for.

[the_watcher]

Sample size doesn't impact the fact that you shouldn't draw causal conclusions from merely correlative data.

[uberman]

Then perhaps you might want to check out some twins studies:

http://davidcard.berkeley.edu/papers/causal_educ_earnings.pd...

and/or

http://www.nytimes.com/1992/08/19/education/twins-study-show...

I'm going to strongly encourage my kids to go to college.

Frankly if you/anyone is not convinced that higher education leads to increased earnings then by all means forgo college and keep your kids at home.

My kids will have plenty of competition for jobs from those with higher degrees from around the world as it is.

[the_watcher]

Pointing out that you should not draw causal conclusions from correlative data does not mean that I would not encourage my kids from going to college. It means that you should not conclude that going to any college causes better outcomes. It also doesn't mean that you should conclude that going to college does not cause better outcomes.

[uberman]

I agree. In fact, I pointed out that the data set I referenced was unlikely to include items like "Parental Social Class" and "IQ" specifically for these reason.

[davvolun]

I think the point is moreso, connecting correlation to causation is really, really, really, really hard, even for simple stuff. And sample size doesn't change that, it may even make it harder (see Simpson's Paradox re: 1964 U.S. Civil Rights Act). I don't know that the_watcher is arguing against college, more noting the underlying assumption and playing devil's advocate.

[ithilglin909]

I would absolutely prefer to send my (hypothetical) children to college for the reasons you've stated here.

But really: are increased earnings due to anything people learn at college or because of its social value? I.e. it's been over-sold and many people will assume you're dumb/couldn't get accepted somewhere if you don't play along.

[whatyoucantsay]

> *My kids will have plenty of competition for jobs

This is the downside of pursuing the same highly commoditized skills being taught to millions of other uni students.

[kimmel]

The 1992 study is basically irrelevant since the job market has changed drastically. There are way more self taught computer people out making money without degrees than in 1992 as just one major factor that has changed in over two decades.

The Berkeley study is missing a lot of data and is dated as well from 1999. I searched for the most common high paying non-degree jobs and found nothing about them. Plumbers, electricians, construction contractors, certain kinds of farmers, etc... Recently on HN it was said that construction works in California were starting at $45 an hour in many places. I know plenty of people who got non-stem degrees that do not make that 10+ years out of college and working professionally in their field the whole time.

Furthermore there is the trend of people getting degrees in fields that will not yield jobs or fields that are over saturated. Take lawyers and this article from 2016 https://www.nytimes.com/2016/06/19/business/dealbook/an-expe... and this one from 2017 https://www.usatoday.com/story/news/2017/06/28/law-schools-h... . There are plenty more articles going back more than a few years about it.

Then there is the issue of colleges and textbooks constantly increasing costs while not increasing the value of the degree itself. In fact in many instances degrees are worth less now since colleges pass students to keep that fire hose of government backed money coming in. There are numerous documentaries on youtube that go into depth of how colleges are ripping people off in many instances.

The main value of college is the ability to network with people so that finding and getting a job becomes easier. The job hunting process is broken and there are always new startups trying to fix it. In high school they pushed college as the solution to everything harder than a drug dealer slinging product. It is easy to find people who got that degree and their job and earning problems are not solved. Going to college no matter what is a dogma.

It is not clear at all that a college education is going to increase your lifetime earning. You have to pick the right field, get your debt settled quickly, live in or move to the right area and conquer the job hunting process and then you might come out ahead. I think high school students should be better educated on how the market works and there are plenty of non-degree jobs that have serious earning potential. Also the US has neglected trade schools as a job training solution while European countries have not.

[uberman]

I'm sure there is nothing that will convince you at this point, however, I'll offer a final set of links.

In 2013 the "conservative" Brookings Institute seemed to agree that there is an economic advantage to a college degree.

see:

https://www.brookings.edu/blog/the-avenue/2013/11/12/the-eco...

and

https://www.brookings.edu/research/where-is-the-best-place-t...

If you feel like they have it wrong and that the last five years has resulted in some sort of sea change then so be it.

Good luck with your kid's high school diploma.

[whatyoucantsay]

> "In 2013 the "conservative" Brookings Institute seemed to agree..."

Setting aside the core causation/correlation problem with your claims about scholastic credentials, it's absurd to call the Brookings institute conservative. It's centre left and, in US election donations, its employees give 97.6% Democrat.

https://www.usnews.com/news/articles/2011/03/03/think-tank-e...