[dahart]

> No they don’t.

I believe you’re wrong on this point, but please elaborate; why do you think that, and what evidence do you have to back it up? Why do you claim that the papers I’ve already linked to aren’t answering this question? Why do you believe that no others exist? Are you talking about causality still, or are you only claiming that the papers I linked didn’t literally talk about a world without college?

Maybe we need to step back and state our assumptions more carefully. The thread above started by @jletienne asked what is the causal vs correlation split for the outcome of a university education, and then @jedberg presumed to clarify that question by asking what if there was no college, which is one of many ways to ask what is the true causal effect of university education on earnings and outcomes, as opposed to correlations that artificially inflate the perceived value of a degree. If people are more likely to go to college because their parents went, then parents deserve some of the credit, and college doesn’t get all of it. If two jobs make use of exactly the same skills, but one asks for a degree and pays more, then the job gets some of the credit too, and college gets even less.

The question ‘what if there was no college’ is simply a way of clarifying the causality, and so I have assumed we’re still talking about the causality question. Are you still talking about the causality question, or are you moving the goal posts on me? Yes it would be ideal to have a world with no college to compare against. The St. Louis Fed paper, since they have no world without college to study, does a causality analysis on the “true” value of college in section III, where they attempt to discount for correlations with some known biases that inflate that perceived value of a degree compared to the world where college never existed.

[rahimnathwani]

I'll paste here GP's question that we're discussing (emphasis mine):

"No, the main question is are the highest paid workers getting that because of college, or did the most driven and smartest go to college, and would have been equally successful had no one gone to college?"

There are a few interesting questions we could ask related to college attendance. The most common ones are:

1. From an individual student's perspective, does going to college have a positive ROI, in a world where the college and employment landscape exists as it does today.

2. If the answer to #1 is yes, how much of this is due to education, and how much is due to pure signalling ("sorting hat") effects?

3. If we were to eliminate all colleges, so that no one has a bachelors' degree, would that reduce the outputs and incomes of the most driven and smartest people.

The papers you cited attempt to address #1 and #2. But neither of those are the question that GP asked (which is #3).

You claim that they do address this question, but are asking me for evidence that they don't. I'm not sure what you're expecting me to do? Go through each and every paragraph in each paper and explain how it doesn't address #3?

It might be easier for you point us to a sentence or paragraph in either paper that envisages a world where no one goes to college.

[dahart]

> neither of those are the question that GP asked

The question of causality is in fact attempting to answer #3, and it seems like you’re failing to understand that. This is exactly what “true causality” means. The question being asked is whether the outputs and incomes of the smartest people are coming from the education, or from the environment where colleges exist. (Implicitly comparing to a world where no colleges exist.) The Fed study and others are trying to answer whether and how much of the outputs and incomes of the smartest people can be assigned to college, and implicitly calculating what the outcomes would be if colleges did not exist.

> It might be easier for you to point us to a sentence or paragraph in either paper that envisages a world where no one goes to college.

Ah, so you want a literal mention of no college. See this is where it becomes clear that you don’t understand the causality question and you didn’t understand the connection between what @jletienne asked and what @jedberg asked. They are not different questions. You think they are, but they aren’t.