Hacker News new | past | comments | ask | show | jobs | submit login

Here is the link to the award (I think): https://www.nsf.gov/awardsearch/showAward?AWD_ID=2245223&His...

Probably half the award is taken up by indirect costs at the university, leaving the remainder for a few PhD students to be funded, money for the devices, and any studies where they are probably paying participants to use the device.

$1.2M doesn't go that far in terms of grants. As far as whether this is a good investment for the government's money, I'm a lot less clear. Given all of the recent car thefts due to TikTok, I assume that influenced NSF and the reviewers.




Half is taken by the university? How does that work?


Renting the space to the departments and other similar things probably.

I wish I was joking, but a friend who works at a Uni was recently complaining of their budget being affected by (mandatory) things like that.


And an evergrowing administration, at least where my mom works. She said they used to take 1/3 not that many years ago, now it's 1/2. In the same period the administration has grown considerably.

A big problem is they don't always get all the money they apply for, and so with the added overhead there's sometimes very little left to do actual science.


Space is limited and coveted. What better way would you propose to allocate it than to give it to who can bring in the money to pay for it? That money goes to building new spaces, which is good for everyone.


> What better way

Open up all federal grants -- especially NSF grants -- to anyone with credentials or experience necessary to PI (so, a PhD or equivalent industry experience). Broaden the reviewer pool so that each panel is at least 51% non-Professor expert citizens.

I can do a LOT of advising and conduct a LOT of research with close to 0% overhead. But most NSF grants are only possible to get if you attach yourself to a university, and at that point it's just not worth the effort. Everyone loses, except for the academic industry, which gets heinously immoral labor laws exceptions so that TT professors and admins can retire-in-place on the taxpayer's dime in their mid 30s.

What possible actual reason does the NSF have for requiring research work to be done at universities?


> I can do a LOT of advising and conduct a LOT of research with close to 0% overhead.

Spoken like someone whose never done research. I’d love to hear what kind of research you think you can do at 0% overhead. No one can do anything at 0% overhead, are you serious?


Have you considered doing an MBA? :)


Sure. Because Universities are all about "bringing in the money" rather than education, research, and similar yeah?


Yeah :(


Consider it rent+taxes to live in the University environment. You get access to libraries. IT infrastructure. Special equipment and space for your research. Lots of competent smart people eager to help you with your project. How much would you pay for that?

I worked at a startup and between the rent for our office and our individual rents, like 80% of the VC money went into landlord pockets.

Also there are plenty of “tax breaks” you can get so that you don’t have to pay so much. Capital expenditures will be taxed at 0%, so you can get your overall rate down significantly.


It varies per university. Looks like for UMich it is 56%. https://orsp.umich.edu/develop-proposal/budget-and-cost-reso...

I believe Stanford is 60%. My university is 54%.

Here is an explanation for the justification of this: https://spo.berkeley.edu/guide/fa.html

It does certainly feel excessive as an academic. I've never seen actual tracking of where the F&A money goes in terms of a quantifiable breakdown, e.g., what fraction goes to university accounting for doing their needed work for supporting a sponsored project, what amount goes toward electricity, etc. Universities always seem to be negotiating with the Federal government to raise the rate. When I was at an institution with a 45% rate, though, it let me stretch my grants a lot further by allowing me to fund more students.


Overhead is required to pay for space, electricity, admin, etc. This is not inherently a problem, but the percentages getting too high is...


Only half. The university bureaucracy is great and all powerful


Happens all the time, nearly everywhere.


Curious, what % of costs of a grant are usually consumed by the UNi - are these to pay for equipment, power, lab time, etc?

Or are they slurping funds for other aspects of the uni's operations?


Usually about half. It's designed to cover the costs of running a university research program spread across all the projects. So a 10MM project provides 10x the funding that a 1MM project provides. Think things like health insurance for the researchers, covering time between projects, lab capital costs, electricity, administration, accounting, even the cost of writing grants.

There's a valid question of if that number can be smaller, but the general concept makes sense.


https://academia.stackexchange.com/questions/25910/what-does...

Overhead includes things like administrative & support staff, equipment depreciation, etc.

It’s often calculated as some percentage of the grant. Then you fill out your timesheet to bill hours to specific grants, so it can all be tracked.


Universities typically take 50%


Honestly good for them. I’m sure they will learn a lot.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: