[ArmchairEcon]

Let's accept your assumption that colleges teach things that are of value in some non-financial sense. Colleges make society better or make people happier. Then we should expect to see an improvement in a King of Bhutan-style Gross National Happiness rather than a reduction of inequality or an increase in GDP.

If Americans aren't, on average, happier than they used to be (adjusted for income), then the conclusion should be "there was no return on the investment in broadening the number of people attending college"

[Retra]

What colleges teach and what they should teach may not be the same. So it's hard to say "if A caused X, we should expect to see X." You can't just assume colleges teach things that are socially valuable because I said they are supposed to.