I believe everyone who taxes ordinary income taxes stipends, although
perhaps there are some states that do not. Certainly Maryland,
Massachusetts and Colorado all tax graduate student stipends. What is
*not* taxed currently is the "tuition reimbursement"; i.e. the $30K+/year
that MIT charges gov't contracts for graduate student tuition (not even
counting benefits & overhead).
I didn't believe Rohit's claim that taxing graduate tuition = a 15% cut in
R&D, so I decided to run some numbers (and please let me know if I've blown
some assumptions):
Assume $15K/year stipend, $30K/year tuition (this is low for MIT these
days, I think, which is $12K+/semester and $7K+ for summer tuition).
Case A: no tax on tuition benefits. Student brings home $15K gross income.
Gets a standard deduction of $4150 and a personal exemption of $2650 in
1997. Net income is $15K - $6,800 = $8,200. Federal tax on $8,200 is
$1,234 (for a single taxpayer, 1996 tables), yielding an after-tax income
of $13,766.
Case B: tuition benefits are taxed as ordinary income. Student brings home
$15K gross income + $30K in tuition = $45K gross income. Net income is
$45K - $6,800 = $38,200. Federal tax on $38,200 is $7,583 (single
taxpayer, 1996 tables), or $6,299 in additional taxes over Case A.
After-tax income is $15K - $7,583 = $7,417.
In order for the student in Case B to end up with $13,766 in after-tax
income, his stipend would have to be about $23,800 (raising gross income
to $53,800, taxes to approx $10,040). That's an increase of about 59% over
the $15K stipend, or a 19.5% increase in the total tuition + stipend amount.
Well, color me surprised. Rohit's 15% figure doesn't look that far out of
whack after all. I guess the real question is whether stipends would rise
enough to offset the increase in taxes. When the NSF stipends became
taxable income in the late 80s they didn't change the stipend amount at
all, so I can't say I'm hopeful.
--bal