The numbers for the MIT living wage calculator are completely wrong for most PhD students. For one, they include thing like social security taxes, which students do not pay. They overestimate medical costs, because most PhD students are young and will not have health complications that drive costs. The rent estimate is based on single-occupancy units, which doesn't make sense because most live with roommates (https://livingwage.mit.edu/resources/Living-Wage-Users-Guide...). Students often qualify for benefits that are not included in the estimate.
It's a popular number to use, but is often quite an overestimate.
This is some nonsense. It's not an overestimate at all.
Grad students do have health issues. I know plenty who do. Actually. I have never had a PhD student who did not have some major health issue in their 5-6 years.
Many grad students have families. They can't live with 6 other people in a unit. They have childcare.
Rent is extremely expensive around MIT. Even a room in a larger unit will easily run you 1.5k or more per month plus utilities.
With 47k in Cambridge even if I shared an apartment and cut expenses I could just scrape by if I didn't have a child. I have no idea how I would make it with one.
Do you disupte any of the things I mentioned? Yes, obviously PhD students with kids and health issues exist, but on average, PhD students are younger, healthier, pay less taxes, qualify for more benefits, and don't have children. The MIT estimate is based on all the above not being true, but that's simply not reflective of the PhD student population.
If you include tuition at the rate undergrads pay it as part of the compensation then they are paid much more than the median wage. Average annual tuition alone for many research universities is higher than the median annual wage.
For the vast majority of the degree, PhDs are not taking any classes (often, they are not even allowed to). They are just workers who either teach or research. The tuition is usually an accounting trick used by colleges/universities for tax and visa sponsorship purposes. Although faculty typically has to cover a PhD's tuition, factoring it in as pay is a little disingenuous.
If tuition was not waived, PhDs would not even apply. The way the PhD is set up in the US is a little byzantine. Other countries (Canada, continental Europe) do have normal employment contracts and benefits.
And this is computer science. If you want to feel bad for someone, grad students in e.g. biology or anthropology are still at or below $20k annually in many places.
It's a popular number to use, but is often quite an overestimate.