[godelski]

Reports like those are not great for people trying to understand data and solve problems. They're for bureaucrats. They're always full of aggregation errors, which is also a common problem many people often get frustrated with from armchair experts. People __should__ be suspicious when reading that the average price paid for college is $2.5k but student debt is almost $2T and the average student debt is nearly $40k. The way the data is presented should cause people to be suspicious because the narrative doesn't make sense. The problem is with the narrative around these numbers, which is how people "lie with statistics" (because math isn't just numbers, data analysis is about stories).

A good way to think about this, and why it should reason that this data at face value isn't useful, is with a rather extreme example. If you took the average income across the globe then you'd conclude that no one could afford college. But that'd be unreasonable because the distribution of incomes is not the same as the distribution of people going to college.

Similarly, in these types of numbers there's a large difference in groups. In fact, contrary to popular belief, those straddled with the most debt are not liberal arts degrees for people with low job prospects but rather high level education of and high prospect jobs. Graduates take out more loans[0]. Medical students take out by far the most[1] -- side note: maybe we should discuss this given the doctor shortage -- followed by PhDs and then masters. Obviously these are in lower numbers though But there are a lot of variables at play here that affect things and it should be rather obvious that you can't treat all these things equally. There are also other issues where private colleges have higher debt burdens and there's also a lot of dubious institutions taking advantage of people (which do we want that happening for people who are specifically trying to become more productive members of society? The most vulnerable are also the "least productive" and even small changes can result in large outcomes).

[0] https://www.brookings.edu/articles/who-owes-all-that-student...

[1] https://wordsrated.com/student-loan-debt-by-degree-statistic...