Are grad students really low paid?

A typical assistantship in a US university grad program pays in the neighborhood of about $1500 a month. Anyone can see that this number is much too low for the time and effort a grad student invests in the research lab or in teaching and would be hardly enough to pull through his/her long years in grad school. Are the grant agencies and universities oblivious to this? No. What we most often miss out is that along with this stipend, the student's tuition is also taken care of. So the actual cost to the professor/department who offers the assistantship, is the stipend (1500*12) + 24 credits a year tuition cost. This is a pretty significant number. In public schools, the tuition could hover around 20K a year. Needless to say, private schools are much more expensive. Once a grad student is done with the course requirement set by the department within a year or two into the program, he/she usually ends up registering for 9 research credits just to stay a full time student. Now, the university typically arrives at a particular cost of X dollars per credit based on various factors that include infrastructure, pay to the faculty, facilities, maintanence, etc.  This number, makes perfect sense for someone who enrolls for a course that is offered by the university. Why then should a research credit also cost the same?

Typically, when a professor gets a grant for pursuing a problem, he has to pay a fraction of that grant to the university that lets him use their resources for conducting research. He then possibly employs a graduate student by offering him an assistantship from that grant money. Towards this assistantship, the professor again ends up paying for the X research credits the student registers for to the university. Isnt this redundant? What is the university's cost towards offering a research credit to a graduate student? Whatever the expense, its already paid for by the fraction that it gets when the grant is awarded to its faculty. 

Why cant the university then charge lesser for its research credits? Just by reducing the cost of the research credit, universities can turn this poor grad student pay scale significantly. The saved money could be diverted into a higher stipend (thereby attracting more students into grad school, eventually increasing the quality of the student pool) or if that is unacceptable (should a section of the academic community decides to keep lab rats poor), pump it back into the project (the research needs money anyway). In any case, its a  win-win situation.