It basically says that earnings depend on the PERSON, not the college.
So what's the point of your project?
Edit to add: I'm not trying to be facetious, but the project seems to be about making decisions to pay for a particular college, which seems at odds with your statement.
I'm not sure I understand your confusion, but let me point out several reasons why it still matters. If college prestige doesn't affect earnings (which the NBER report shows), and your goal is to maximize NPV of college, then you want to go to an inexpensive college. This helps you compare colleges by cost. Furthermore, if you have multiple majors that you are considering that suit your interests, then this helps you assess which major might be more lucrative. It then combines these two factors to give you an estimate of your financial outcome. Most college ranking sites only consider the college costs in isolation, when major choice is at least as important. On top of all that, it gives you a baseline comparison to 2-year degrees (on average) and high school only earnings to give some perspective by making a relative comparison against other paths a prospective college student might consider.
But your site says it's an analysis of "your prospective college and career path versus alternative," but now you're saying that that just means the cost of that college (not the job and salary development), and you're saying that we can ignore that side of the equation on the basis of a study based on admissions decisions from 35 years ago. Have there been changes in college admissions in the past 35 years? In the economy generally, we read regularly that vast portions of all income gain in that time period have gone to a small group---you're confident that's not relevant?
And regardless of whether it's the student or his or her choice of college that primarily shapes later income, it still doesn't make sense to me to generate a report that reflects an income projection that has nothing to do with either the student or the college and instead just reflects an overall average.
I'm definitely not saying that. I'm not sure if you are only skimming my text, but I said quite clearly that "It then combines these two factors to give you an estimate of your financial outcome." The two factors being college expenses, and career earnings - indicating that it is taking both college choice/cost and career/major into account.
So let me be more clear in case I am not. You originally indicated that you agreed with TLV that the report may not be taking into account a variance in earnings due to college choice (e.g. MIT grads earn more). A valid concern which I also shared while developing the site until I found the NBER report. What the NBER report says is that variance in earnings for a given major is not due to college attended, but that the variance in earnings is due to an individuals characteristics (whatever those may be). Mathematically, that is: Pr(salary | major, person, college) = Pr( salary | major, person) and thus Var( salary | major, person, college) = Var( salary | major, person). I.e. college attended is an irrelevant parameter in predicting the variance in earnings given a person (per the NBER findings).
I suppose you can argue with the NBER report findings. Fine. But if it is correct,
you have to realize how incredibly important it is. The common narrative is that you have to get into a "top tier" or Ivy League school to make higher than average salary for your field, but the NBER report contradicts that, saying that graduates of top tier schools earn more (on average, than the average for a major) because they attract people that have better earning potential, for whatever reason (work ethic, intelligence, etc.), but that the individuals higher-than-average earning potential is present regardless of what college he/she attends. That's huge because it means if you are really smart/hard working/lucky/whatever you don't need to pay for a top tier degree to have that higher than average salary; you will still get that higher salary, on average, if you go to an inexpensive state school.
As to "we read regularly that vast portions of all income gain in that time period have gone to a small group---you're confident that's not relevant?" It's possible, although it seems that's largely due to wealth concentration leading to more wealth concentration, and doesn't seem related to college choice. Frankly, there's a lot of things I can't be confident are not relevant. GPS coordinates of person's birthplace? Astrological sign the sun was in when the college was founded? Years a major has been available? Frankly, if I could model every possible factor and had the statistical distributions on those factors, I would be doing this on Wall Street and making so much cash that buying a Lambo would be mustachian. I would probably also hate my life, but that's another thing.
But just so I'm really clear on what's going on, let me give you a simplified version of the model the site runs through to generate the NPV in the report:
NPV = f(College Expenses, Career Earnings, discount rate) and...
College Expenses = f( college tuition and other expenses, increase rates, years in college)
Career Earnings = f( major, annual raise rate, years worked)
Keep in mind all monetary values in the calculation are brought to present value through the discount rate.
Could I improve earning estimates if I knew more about the individual? Yes, by asking things like IQ score, socioeconomic background, etc., but it's not really feasible, and people tend to inflate those factors when self-reporting. This is why Section 8 is so important. Specifically 8.3 - Lifetime Earnings, which shows how your NPV will change if you're an overachiever/underachiever, which I leave the user to judge for themselves.
Hope that is clear enough.