Productivity isn’t the metric that determines whether the government recoups student loan defaults, income is. And the income of graduates is still multiples that of non-graduates, as shown by the data in this article.
Your hypothetical isn’t the way things are actually playing out, so I don’t feel like it needs to be argued. Degree holders are generating value on average, so the economy does, in fact, have more money as a result of education. And the government also makes money on it’s investment in education on top of that.
Generating value is synonymous with growing the economy and being more productive. You can't inflate incomes without either increasing productivity or printing money. If degree holders command higher incomes without offering higher productivity, it comes at the expense of non-degree holders earning lower wages than they would have without the degree holders around.
In my opinion, much of the income premium comes from employers using degrees as a filter (similar to a standardized test). Doing well on the SAT doesn't make you more productive, but does give you a leg up over your peers.
I’m not really sure how to answer your first part, I’m not trying to talk past anyone, but I feel like people are responding to my point that the income premium is what matters to the student loan program by repeating the conclusion of the paper that the wealth premium is going to zero.
The income premium and the wealth premium in this paper are two different things, and even though people with degrees might end up having net zero additional wealth accumulation at the end of their lives, that doesn’t mean the student loan program is also net zero, or that people with and without degrees have the same lifestyle.
The student loan program gets to make it’s money back before degree holders get to save the money, so the income premium matters to the government more than the wealth premium, when figuring out whether forgiving loans is worth it or not.
The example of patents looks a lot more like "people with advanced degrees are more likely to be in jobs where filing for patents is encouraged by the employer" than it does like "people with advanced degrees invent more". To be honest, the fact that our current patent situation is so obviously a net negative might be skewing my judgement on this, but if you thought that bringing up patents was a way to make it sound convincing that college degrees enhance productivity, well it doesn't for me anyway. I even have a patent, or my employer got one from my work anyway, and I still think our patent system is value-destructive.
Maybe the government part is confusing here. Imagine student loans coming strictly from for-profit banks. It is possible for the banks to make their money back, while students lose money, right? This happens to restaurant owners and poor families all the time. It doesn’t matter if the loans stimulate the economy, or if a percentage of loans default, the bank still makes it’s money.
Anyway, I’ll play along with the other arguments, just for fun. So by what metric are you declaring that productivity is not growing per employee? The US per-capita GDP has been growing steadily. How do you back up your claim that productivity hasn’t gone up?
Even if it were true, I don’t think it’s clear or true that you can declare the economy as not having benefitted from education in the past. The paper we’re commenting on has data that shows a net positive economic value in the past, even though the wealth premium is disappearing. And BTW do you really believe that - that the education system has done zero economically? Is that what you’re implying?
People with advanced degrees do invent more, in part because people with advanced degrees are more likely to be in jobs where invention and patents are encouraged, yes. Also because people who are prone to inventing are more likely to choose to do a PhD. How does this demonstrate that people would be inventing more if they chose not to go to college? People can already choose not to go to college, and choose to invent, and by and large that is not happening - what’s really stopping all the non-degree holders from inventing right now? You’re saying it’s just jobs keeping inventors out? There’s no education requirement on submitting a patent, so if college does nothing, why aren’t high school graduates already the majority of applications?
Of course, degrees aren't zero value, but the subsidies for education (student loans) could still be distorting supply in a way that results in a net loss for the government. For example, a self-taught engineer or entrepreneur spends years in school instead of inventing or starting businesses. This is a poor allocation of investment and human capital which results in a net economic loss (including a loss for the government).
To put it another way, earnings are correlated with an individual's talent and drive. Currently, many of those people end up getting degrees. It's a correlation, it's not the degrees that are causing the jump in earnings, it's the individual's skills.
To address your separate point directly, there are studies that attempt to directly answer the question you’re asking. For example: http://conference.iza.org/conference_files/CoNoCoSk2011/gens...
However, your conclusion isn’t shared by the authors. They show a positive return on college degrees, it’s just smaller than the unadjusted difference in income.
One probable factor for this is cultural bias toward giving jobs to people with degrees, and paying employees more who have degrees. Individual skills alone do not explain the difference in earnings, so despite the fact that talent correllates with income, the correlation is not 1, and thus it is not likely that the same people earn equal or greater without a degree than with. It is likely that the same person without a degree will earn less than with a degree, on average.
In fact, the labor productivity growth rate does not seem to have significantly changed at all, even as the percentage which is college educated has gone up a lot: https://stats.oecd.org/Index.aspx?DataSetCode=PDYGTH
To address your point directly, I don’t personally believe that non-grad programmers are as economically productive as grads on average. I would totally believe that non-grads are as productive at building UI widgets and CRUD apps and stuff like that. However, inventions and patents are absolutely dominated by people with postgraduate degrees. It’s not a matter of hourly output, it’s a matter of yearly compound interest, and the compound interest in computer programming is generated by research. It is overwhelmingly PhDs who are developing neural networks and AI methods right now, overwhelmingly postgraduates who are the architects of new hardware and of new software methods. These are the things making the computer industry grow larger in fundamental ways.