First, the USNews article you cite says the average is $57k. The $100k number is the 75th percentile number.
Second, the article you cite is pretty freely mixing "graduate students" and "graduate and professional degrees", to the extent that it's not clear whether the average numbers it cites are just for the former, or for the latter. It obviously makes a big difference, because while law school and medical school have their own problems they don't have the (very real) problems you describe PhD programs as having.
So I looked at the actual report your linked article is citing. The $57k number is the _median_ (not average!) debt across all graduate and professional students. Same for the $100k number: 75th percentile across all graduate and professional students. See page 1 of the report.
Page 3 of the report cites some "typical" (I assume they mean "median", but they don't define it) numbers for various graduate degree debts, including law and medicine, but conveniently leaves out the number for "PhDs". Those make up about 23% of all graduate degrees, according to the chart on that page, by the way, so your typical "graduate or professional student" is not a PhD candidate, but is aiming for an MD, JD, or Master's degree, all of which have _quite_ different funding models from PhDs.
All the tables on page 12 and following conveniently exclude PhDs as well.
So this report tells us pretty much nothing about PhD debt. The law and medicine numbers inflate everything involved, obviously, and most of the rest are masters degrees of various sorts. All PhDs could have a debt of 0 and still get the reported median and 75th percentile numbers.
OK, so how this works in practice (or at least did 10 years ago) at the universtity of Chicago, while I wad doing my PhD there.... Grad students in the _sciences_ generally did not take on debt at all: their tuition was covered, and they were paid a stipend that was enough to live on reasonably, in return for the teaching and whatnot that they did. Grad students in the _humanities_ were an entirely different story. So even within the PhD bucket it really depended on the field of study. The number of students who took on debt and had "lab time" of any sort was quite close to 0, if not exactly 0.
Now there are real problems in PhD programs, including in the sciences, and grad students and especially postdocs _are_ underpaid in various ways. But you're not having science PhDs with $100k in grad school debt, typically.
First, the USNews article you cite says the average is $57k. The $100k number is the 75th percentile number.
Second, the article you cite is pretty freely mixing "graduate students" and "graduate and professional degrees", to the extent that it's not clear whether the average numbers it cites are just for the former, or for the latter. It obviously makes a big difference, because while law school and medical school have their own problems they don't have the (very real) problems you describe PhD programs as having.
So I looked at the actual report your linked article is citing. The $57k number is the _median_ (not average!) debt across all graduate and professional students. Same for the $100k number: 75th percentile across all graduate and professional students. See page 1 of the report.
Page 3 of the report cites some "typical" (I assume they mean "median", but they don't define it) numbers for various graduate degree debts, including law and medicine, but conveniently leaves out the number for "PhDs". Those make up about 23% of all graduate degrees, according to the chart on that page, by the way, so your typical "graduate or professional student" is not a PhD candidate, but is aiming for an MD, JD, or Master's degree, all of which have _quite_ different funding models from PhDs.
All the tables on page 12 and following conveniently exclude PhDs as well.
So this report tells us pretty much nothing about PhD debt. The law and medicine numbers inflate everything involved, obviously, and most of the rest are masters degrees of various sorts. All PhDs could have a debt of 0 and still get the reported median and 75th percentile numbers.
OK, so how this works in practice (or at least did 10 years ago) at the universtity of Chicago, while I wad doing my PhD there.... Grad students in the _sciences_ generally did not take on debt at all: their tuition was covered, and they were paid a stipend that was enough to live on reasonably, in return for the teaching and whatnot that they did. Grad students in the _humanities_ were an entirely different story. So even within the PhD bucket it really depended on the field of study. The number of students who took on debt and had "lab time" of any sort was quite close to 0, if not exactly 0.
Now there are real problems in PhD programs, including in the sciences, and grad students and especially postdocs _are_ underpaid in various ways. But you're not having science PhDs with $100k in grad school debt, typically.