[the_snooze]

Research overhead doesn't pay for that stuff. It pays for research-relevant infrastructure and staff. If a PI buys equipment, it's a waste to spend their own very expensive time to ensure purchase-tracking compliance. So you hire an inventory person. But it doesn't make sense to have one inventory person per research group. So you hire one to serve multiple projects and ensure their continued employment by averaging out the costs across those funding lines. That's the overhead cost.

[transcriptase]

“Purchase compliance” with what? Arbitrary policies set by… university admin?

Full disclosure I’ve dealt with this all first hand. The vast majority of universities contribute next to nothing to research programs except the space, utilities, and IT. And those costs do not need to be a % of incoming grant money.

And the vast majority do just take that cut and use it as slush fund general revenue.

[eesmith]

You are right that it does not "need to be a %" - it could be a la carte, for example.

Our physics group had access to the shared LN2 tank. It was no fee, so if you needed a few liters you could just grab it. That could be replaced with a debit system which goes directly to the account. (Hmm, and who decides how much to charge?)

We got hazmat training, and the environmental safety officer was paid by the university through overhead. That could instead be billed directly to the grant. (Hmm, and who decides how much to charge, and how much training is needed?)

When we installed a projector on the ceiling we were able to get building services to install the mount through the air plenum. That service could instead be billed directly against the grant. (Hmm, and who decides how much to charge?)

We could book time in the sound studio to record voice-overs for one of our projects. That was free, but could be billed directly against the grant. (Hmm, and who decides how much to charge?)

Some relevant materials, believe it or not, are not yet digitized, but are available in the library stacks. I could just walk there and read it, because library support was part of the overhead, but you're right - it doesn't need to be that way. The library could charge per entry and book. (Hmm, and who decides how much to charge?)

I'm sure other departments have their own needs, like maintaining the herds for the ag school, or the instruments for the music department.

I don't know how you could have dealt with this all first hand and not see just how much more the university provides than "space, utilities, and IT."

[the_snooze]

Funding sources themselves, especially with capital expenditures for purchases $10K or higher. They want to make sure PIs are spending money on actual program equipment that's being used, and not just turning around and selling it on Ebay to line their pockets.

I've dealt with this first hand because I'm a PI myself. I hate having to waste my valuable time dealing with the inventory person every couple months, but we do it because our funding sources require that kind of accountability.

[transcriptase]

That’s interesting. In Canada the issue is that funding sources don’t care if you’re somewhat irresponsible with capital expenditures (though not to the point of personal enrichment), but are loathe to give you any money for personnel.

[jeffbee]

> I’ve dealt with this all first hand ... In Canada ...

Consider sitting this one out.

[secabeen]

Sadly, many of those arbitrary policies are set by ... The Federal Government! As an example, NIH just added a huge new cybersecurity requirement for a bunch of their datasets.

Space, utilities and IT can be significant costs. There are also the standard costs of personnel for the people working under these grants (HR, Accounting, Purchasing, etc.) Add onto that the significant compliance requirements.

When the government hires a consultant, they usually pay 2-3x the salary of the consultant in per-hour costs. This rule limits them to 1.15x the salary of the employee.

[citrate05]

Compliance with policies set by Congress and the relevant funding agencies, which establish standards for things like safeguarding the private health data and safety of study participants, mandating that de-identified versions of data be made available to other labs so they can both verify and build on the research, biosecurity, preventing foreign interference, lab animal welfare, hazardous waste disposal… and ironically, making sure that money is being spent appropriately by PIs.

See e.g. https://www.cogr.edu/sites/default/files/1991%20Excel%20Data...

[robwwilliams]

Arbitrary policies often set by the federal and state governments. Yes sure, there is some policy at the institutional level but the intent is usually good—-to prevent fraud and abuse and get a good price. Certainly not to slow down research.