19 ms·

When I was a PhD student in Computer Science, I wanted to fund my research without depending on other people. So I wrote, and won, a grant for $650,000. The U

by toomim 2y ago

When I was a PhD student in Computer Science, I wanted to fund my research without depending on other people. So I wrote, and won, a grant for $650,000. The University of Washington took 55% off the top in "indirect costs." Then of the direct costs, a bunch more taxes went to the department, the lab, etc. I ended up getting about $40,000 total from that grant to fund my actual research. I was told that this is just how the game works. The Lesson: There is a huge amount of waste in academic grants. The university does not need 55% indirect costs to fund my research. They gave me a laptop and an office to work in. But the laptop actually came out of direct costs. And as for the office... well, I tended not to use it. Coffeeshops were nicer places to work. So that 55% — $357,500 — was not necessary for research. It disappeared in an administrative hierarchy that kept growing. I talked to professors from 20 years before, and they said that the administration had grown 3x in their time being there. They built their reputation as scientists when everything was smaller and cheaper. Now all that reputation is generating more money, which is getting gobbled up by an administration that wasn't needed when they produced their original successes.

KennyBlanken 2y ago
"I got an office but I decided I'd rather be in a coffee shop" doesn't mean you should get a refund... I love it when people misrepresent things to people who don't know better. This happens every time research funding gets mentioned on HN, by people looking to rile up HNers who absolutely love hearing that their prejudices against the government and scientists are true and academia is robbing the dear taxpayer blind, like it's all some big shady scam. Direct costs are things that are directly and solely related to or used for the grant. Labor. Equipment. Supplies. For the research funded by the grant, within reason (nobody accounts for or cares about, say, a lab mate asking to use some of a chemical, or one of your DVD-R disks. If you have plenty of room in that incubator that was paid for with your grant and a labmate or someone in the department could really use a corner of it to try something on their experiment, or their incubator dies and you've got room? You say yes. Captain NIH does not descend and go HALT SCIENTIST! THAT EQUIPMENT WAS PAID FOR OUT OF DIRECT COSTS FOR YOUR GRANT, NOT THEIRS!" Indirect costs are all the costs necessary for "keeping the lights on." Indirect costs are like taxes because everyone's grants are "taxed" and that pays for all the things everyone uses. Maintaining the buildings, paying the electric bill, paying the salaries of the IT staff, paying for network equipment, servers, storage, UPSs. Indirect grant costs pay for the trash (and hazardous waste) to get collected, for the parking lot to get plowed in the winter, for the buildings to be cooled in the summer and heated in the winter. In a lab it pays for stuff like centralized purified water, central gas supplies, and so on. You don't get to cry about how your work doesn't need all that lab stuff those bio researchers need, or that big storage cluster so you couldn't have to pay for it. Just like you don't get to cry about how the road you live off of is very short so your taxes should be lower because the town has to pay less to pave and plow it, and since you don't have any kids your taxes should be lower still because you're not using the schools. And really, your house didn't catch on fire (and you're very careful about fire safety) so why do you have to pay for the fire department? etc. Indirect costs are negotiated between schools and the organizations that award grants. A school doesn't get to make up whatever they want. It's known by everyone involved, well ahead of time. Nobody walked into Toomim's office and said "ooo, nice shiny $650k grant you got there, Michael Toomim. We're gonna just...take...about half of that. YOINK! SUCKER! Where does everyone want to go next weekend? Bermuda, or Italy? WHY NOT BOTH!" Further, indirect costs are accounted for when grants are awarded. The awarding organization didn't think his grant proposal was worth $650k and then oooo, the Big Mean School Administration came and took half his stuff and poor Michael Toomim was left with half the money he thought he needed...frowny face :( The organization thought the work proposed in the grant application would require $300k in direct costs. They looked up what his school uses for indirect costs, and "wrote the check" for $650k. Other grant organizations award an amount and then an additional amount is added equal to whatever the school charges for indirect costs. Everyone knows this, grad students included. Nobody counts indirect costs as part of their grant. Ever. When they say "I got a million dollar grant!" they mean "a million dollars plus indirect costs charged by the organization." Third, there is no fucking way whatsoever that the school took any additional money from the grant for anything except direct costs - stuff that was used solely and exclusively for that grant. Nothing in the grant was yoinked "by the lab" or "by the department." That's called "double dipping." Every org that gives grants prohibits it, and if you're caught doing it, it is a big fucking deal.
bakuninsbart 2y ago
Just yesterday I watched a video by a disgruntled former academic physicist about the state of modern physics research. [0] Not sure of its credence, I don't really understand physics. In the fields I do have a bit of understanding, there does seem to be an amazing amount of fluff and bloat, but also still a significant amount of good and useful knowledge creation. The bloat isn't really surprising: Largely inoculated from market forces, the way modern research is conducted has enormous amounts of inefficiencies, which in turn creates plenty of ways to game the system. Without university accreditation it is very hard to publish research, get grants or invites to conferences, so there is little reason for universities to take less of your grant. Why should journals not be super profitable? It is one of the very few "objective" metrics of success in academia, Nature could probably introduce expensive publishing fees tomorrow, and people would still fall in line. And universities don't really have to adapt, because going to one is basically required to get a good paying job. I don't know how to fix that, it seems like a very hard problem. Just shutting everything down and leaving it to the private market as the Trump administration is aiming to do seems overkill, but sometimes it might also be necessary to break a stagnating system before rebuilding it better. [0] https://www.youtube.com/watch?v=shFUDPqVmTg https://www.youtube.com/watch?v=shFUDPqVmTg
czzr 2y ago
Do you think medical researchers also do all their work in coffee shops?
jltsiren 2y ago
You have misunderstood a number of things. Overheads are expressed as a fraction of direct costs, not of the total grant. If the total grant was $650k, 55% overhead would be $230k. If the direct costs were $650k, the total grant was $1M. Universities rarely have significant amounts of money they can use freely. Most of the time, if they want to do something (such as training a PhD), someone must pay specifically for that. And if the PhD student needs a supervisor, someone must pay for the time the supervisor uses supervising the student. Your grant was not supposed to give you a huge pile of loose money to use as you see fit. It was supposed to pay your stipend and tuition (and benefits, if there were any). And give you a small amount for research expenses on top of that. You gained independence (by not being dependent on your supervisor's funding) and credibility (by winning a competitive grant). You may have also gained the ability to continue your PhD (if your supervisor was about to run out of money). But often the biggest winner with grants like that is the supervisor, who can now use their grant money to hire another PhD student.
- toomim 2y ago
  You aren't clear with what I supposedly "misunderstood." The entire grant was $650k. 55%, or $357k were used for indirect costs. I don't know how you multiplied 650 * .55 and got 230, or why you're thinking about a $1M grant.
  - pmags 2y ago
    That's not how indirects are calculated. Assuming the entire award was subject to the 55% indirect rate (not always true), and the total award was $650K, then the direct award amount was about $420K and the indirect amount was ~$230K. $420K direct $420K * 0.55 = ~$230K indirect Total: $650K
pmags 2y ago
Everything about this sounds highly unusual: 1. Universities do not typically allow graduate students to be PIs on grants. This is for liability reasons. Even for postdocs it would be highly unusual to have full, solo PI status on a grant. 2. Federal funding agencies, and the panels that review grants, would be highly dubious about a graduate student as the sole PI on a grant. For example, NSF's Proposal & Award Policies & Procedures Guide (PAPPG) (NSF 24-1; https://www.nsf.gov/policies/pappg/24-1/ch-1-pre-submission#ch1E https://www.nsf.gov/policies/pappg/24-1/ch-1-pre-submission#...) explicitly states: "Graduate students are not encouraged to submit research proposals but should arrange to serve as research assistants to faculty members." NIH is even more constrained w/respect to graduate students. Even pre-doctoral fellowships (F31) at NIH to support a specific graduate student are submitted by a faculty member. I can't speak to DOE or DARPA policies off the top of my head on this point. 3. Could you provide a grant/award number to lookup on NSF Award Search (https://www.nsf.gov/awardsearch https://www.nsf.gov/awardsearch), NIH Reporter (https://reporter.nih.gov/ https://reporter.nih.gov/) or the award search page of whatever federal agency awarded this grant?
- toomim 2y ago
  You're misunderstanding. I wasn't the PI, and never said I was. My PhD advisor was the PI. I wrote the grant as a graduate student, and did the research. It was an NSF grant, with a title on something about Social Computing Systems, in like 2010 or 2011 or so. You can search for it if you want.
  - rsfern 2y ago
    So what did the 40k for your research cover? It sounds like you are probably discounting the cost of your own salary as a graduate student supported by the grant, which is usually the bulk of the direct cost for NSF grants like this and probably runs 100k/year, with the grant covering maybe 3 years of your time.
  - pmags 2y ago
    Perhaps this award? James Landay (Principal Investigator) Claus Portner (Co-Principal Investigator) Interaction Economics: Instruments that Measure Social-Computational Systems https://www.nsf.gov/awardsearch/showAward?AWD_ID=1110965 https://www.nsf.gov/awardsearch/showAward?AWD_ID=1110965 There's a single publication listed as resulting from this award (NB: I'd guess there's probably additional papers stemming from this grant; NSF award search only lists what has been published at the time of the annual report for the FY referenced). That publication includes 4 co-authors: Michael Toomim, Travis Kriplean, Claus Portner, and James A. Landay "Utility of Human-Computer Interactions: Toward a Science of Preference Measurement" ACM Conference on Human Factors in Computing Systems (CHI) , v.May , 2011 This doesn't sound like a "research without depending on other people" type situation, unless you're suggesting your co-authors didn't contribute meaningfully to the research. We can't pull up the detailed budget, but let's do some quick math for what such a grant would typically fund for a CS type project. Assuming a 3-year grant ($140K/year) the directs would have looked like this: * 2 graduate students - 12 months of stipend, fringe, and tuition/fees. This will be the largest portion of the grant; Total: ~$95K/year (Note: 2010 grad student support costs; would be lot more now): * 2 PIs -- probably 0.5 or 1 month summer salary ([1] see below) for each PI each plus fringe; Total: $40K/year * A small travel budget to pay for partipants to attend a conference or meeting (like the ACM conference this paper was presented at); Total: $5K/year [1]: Faculty in most schools of arts and sciences schools at major universities are only guaranteed a 9-month salary (paid out over 12 months). Note this can differ even within a University. For example, A&S faculty might have one standard; the School of Medicine another; etc.
amazingman 2y ago
Downthread it becomes clear that you are dramatically distorting the truth. What is your motivation for this?

by toomim 2y ago

KennyBlanken 2y ago

"I got an office but I decided I'd rather be in a coffee shop" doesn't mean you should get a refund... I love it when people misrepresent things to people who don't know better. This happens every time research funding gets mentioned on HN, by people looking to rile up HNers who absolutely love hearing that their prejudices against the government and scientists are true and academia is robbing the dear taxpayer blind, like it's all some big shady scam. Direct costs are things that are directly and solely related to or used for the grant. Labor. Equipment. Supplies. For the research funded by the grant, within reason (nobody accounts for or cares about, say, a lab mate asking to use some of a chemical, or one of your DVD-R disks. If you have plenty of room in that incubator that was paid for with your grant and a labmate or someone in the department could really use a corner of it to try something on their experiment, or their incubator dies and you've got room? You say yes. Captain NIH does not descend and go HALT SCIENTIST! THAT EQUIPMENT WAS PAID FOR OUT OF DIRECT COSTS FOR YOUR GRANT, NOT THEIRS!" Indirect costs are all the costs necessary for "keeping the lights on." Indirect costs are like taxes because everyone's grants are "taxed" and that pays for all the things everyone uses. Maintaining the buildings, paying the electric bill, paying the salaries of the IT staff, paying for network equipment, servers, storage, UPSs. Indirect grant costs pay for the trash (and hazardous waste) to get collected, for the parking lot to get plowed in the winter, for the buildings to be cooled in the summer and heated in the winter. In a lab it pays for stuff like centralized purified water, central gas supplies, and so on. You don't get to cry about how your work doesn't need all that lab stuff those bio researchers need, or that big storage cluster so you couldn't have to pay for it. Just like you don't get to cry about how the road you live off of is very short so your taxes should be lower because the town has to pay less to pave and plow it, and since you don't have any kids your taxes should be lower still because you're not using the schools. And really, your house didn't catch on fire (and you're very careful about fire safety) so why do you have to pay for the fire department? etc. Indirect costs are negotiated between schools and the organizations that award grants. A school doesn't get to make up whatever they want. It's known by everyone involved, well ahead of time. Nobody walked into Toomim's office and said "ooo, nice shiny $650k grant you got there, Michael Toomim. We're gonna just...take...about half of that. YOINK! SUCKER! Where does everyone want to go next weekend? Bermuda, or Italy? WHY NOT BOTH!" Further, indirect costs are accounted for when grants are awarded. The awarding organization didn't think his grant proposal was worth $650k and then oooo, the Big Mean School Administration came and took half his stuff and poor Michael Toomim was left with half the money he thought he needed...frowny face :( The organization thought the work proposed in the grant application would require $300k in direct costs. They looked up what his school uses for indirect costs, and "wrote the check" for $650k. Other grant organizations award an amount and then an additional amount is added equal to whatever the school charges for indirect costs. Everyone knows this, grad students included. Nobody counts indirect costs as part of their grant. Ever. When they say "I got a million dollar grant!" they mean "a million dollars plus indirect costs charged by the organization." Third, there is no fucking way whatsoever that the school took any additional money from the grant for anything except direct costs - stuff that was used solely and exclusively for that grant. Nothing in the grant was yoinked "by the lab" or "by the department." That's called "double dipping." Every org that gives grants prohibits it, and if you're caught doing it, it is a big fucking deal.

bakuninsbart 2y ago

Just yesterday I watched a video by a disgruntled former academic physicist about the state of modern physics research. [0] Not sure of its credence, I don't really understand physics. In the fields I do have a bit of understanding, there does seem to be an amazing amount of fluff and bloat, but also still a significant amount of good and useful knowledge creation. The bloat isn't really surprising: Largely inoculated from market forces, the way modern research is conducted has enormous amounts of inefficiencies, which in turn creates plenty of ways to game the system. Without university accreditation it is very hard to publish research, get grants or invites to conferences, so there is little reason for universities to take less of your grant. Why should journals not be super profitable? It is one of the very few "objective" metrics of success in academia, Nature could probably introduce expensive publishing fees tomorrow, and people would still fall in line. And universities don't really have to adapt, because going to one is basically required to get a good paying job. I don't know how to fix that, it seems like a very hard problem. Just shutting everything down and leaving it to the private market as the Trump administration is aiming to do seems overkill, but sometimes it might also be necessary to break a stagnating system before rebuilding it better. [0] https://www.youtube.com/watch?v=shFUDPqVmTg https://www.youtube.com/watch?v=shFUDPqVmTg

czzr 2y ago

Do you think medical researchers also do all their work in coffee shops?

jltsiren 2y ago

You have misunderstood a number of things. Overheads are expressed as a fraction of direct costs, not of the total grant. If the total grant was $650k, 55% overhead would be $230k. If the direct costs were $650k, the total grant was $1M. Universities rarely have significant amounts of money they can use freely. Most of the time, if they want to do something (such as training a PhD), someone must pay specifically for that. And if the PhD student needs a supervisor, someone must pay for the time the supervisor uses supervising the student. Your grant was not supposed to give you a huge pile of loose money to use as you see fit. It was supposed to pay your stipend and tuition (and benefits, if there were any). And give you a small amount for research expenses on top of that. You gained independence (by not being dependent on your supervisor's funding) and credibility (by winning a competitive grant). You may have also gained the ability to continue your PhD (if your supervisor was about to run out of money). But often the biggest winner with grants like that is the supervisor, who can now use their grant money to hire another PhD student.

toomim 2y ago
You aren't clear with what I supposedly "misunderstood." The entire grant was $650k. 55%, or $357k were used for indirect costs. I don't know how you multiplied 650 * .55 and got 230, or why you're thinking about a $1M grant.
- pmags 2y ago
  That's not how indirects are calculated. Assuming the entire award was subject to the 55% indirect rate (not always true), and the total award was $650K, then the direct award amount was about $420K and the indirect amount was ~$230K. $420K direct $420K * 0.55 = ~$230K indirect Total: $650K

pmags 2y ago

Everything about this sounds highly unusual: 1. Universities do not typically allow graduate students to be PIs on grants. This is for liability reasons. Even for postdocs it would be highly unusual to have full, solo PI status on a grant. 2. Federal funding agencies, and the panels that review grants, would be highly dubious about a graduate student as the sole PI on a grant. For example, NSF's Proposal & Award Policies & Procedures Guide (PAPPG) (NSF 24-1; https://www.nsf.gov/policies/pappg/24-1/ch-1-pre-submission#ch1E https://www.nsf.gov/policies/pappg/24-1/ch-1-pre-submission#...) explicitly states: "Graduate students are not encouraged to submit research proposals but should arrange to serve as research assistants to faculty members." NIH is even more constrained w/respect to graduate students. Even pre-doctoral fellowships (F31) at NIH to support a specific graduate student are submitted by a faculty member. I can't speak to DOE or DARPA policies off the top of my head on this point. 3. Could you provide a grant/award number to lookup on NSF Award Search (https://www.nsf.gov/awardsearch https://www.nsf.gov/awardsearch), NIH Reporter (https://reporter.nih.gov/ https://reporter.nih.gov/) or the award search page of whatever federal agency awarded this grant?

toomim 2y ago
You're misunderstanding. I wasn't the PI, and never said I was. My PhD advisor was the PI. I wrote the grant as a graduate student, and did the research. It was an NSF grant, with a title on something about Social Computing Systems, in like 2010 or 2011 or so. You can search for it if you want.
- rsfern 2y ago
  So what did the 40k for your research cover? It sounds like you are probably discounting the cost of your own salary as a graduate student supported by the grant, which is usually the bulk of the direct cost for NSF grants like this and probably runs 100k/year, with the grant covering maybe 3 years of your time.
- pmags 2y ago
  Perhaps this award? James Landay (Principal Investigator) Claus Portner (Co-Principal Investigator) Interaction Economics: Instruments that Measure Social-Computational Systems https://www.nsf.gov/awardsearch/showAward?AWD_ID=1110965 https://www.nsf.gov/awardsearch/showAward?AWD_ID=1110965 There's a single publication listed as resulting from this award (NB: I'd guess there's probably additional papers stemming from this grant; NSF award search only lists what has been published at the time of the annual report for the FY referenced). That publication includes 4 co-authors: Michael Toomim, Travis Kriplean, Claus Portner, and James A. Landay "Utility of Human-Computer Interactions: Toward a Science of Preference Measurement" ACM Conference on Human Factors in Computing Systems (CHI) , v.May , 2011 This doesn't sound like a "research without depending on other people" type situation, unless you're suggesting your co-authors didn't contribute meaningfully to the research. We can't pull up the detailed budget, but let's do some quick math for what such a grant would typically fund for a CS type project. Assuming a 3-year grant ($140K/year) the directs would have looked like this: * 2 graduate students - 12 months of stipend, fringe, and tuition/fees. This will be the largest portion of the grant; Total: ~$95K/year (Note: 2010 grad student support costs; would be lot more now): * 2 PIs -- probably 0.5 or 1 month summer salary ([1] see below) for each PI each plus fringe; Total: $40K/year * A small travel budget to pay for partipants to attend a conference or meeting (like the ACM conference this paper was presented at); Total: $5K/year [1]: Faculty in most schools of arts and sciences schools at major universities are only guaranteed a 9-month salary (paid out over 12 months). Note this can differ even within a University. For example, A&S faculty might have one standard; the School of Medicine another; etc.

amazingman 2y ago

Downthread it becomes clear that you are dramatically distorting the truth. What is your motivation for this?