3 ms·

Antedote: When I was a PhD student at the University of British Columbia in the Department of Computer Science research grant proposals included budgets for ov

Antedote: When I was a PhD student at the University of British Columbia in the Department of Computer Science research grant proposals included budgets for overhead. This was often, if I recall correctly, 10% of the overall equipment and salary budget. This was deemed a tax collected by the department and used to: - pay IT staff salaries - pay IT hardware and service costs (storage, communication) - etc Other costs on the research grant would include: - hardware purchases: personal computers, specialized equipment, compute and storage servers - graduate student and postdoctoral student salaries - travel costs to conferences for students - consumables (if any) Again, in Canada, there are various types of research grants. For example, the "Research Tools and Instruments grants program" will pay the entire cost of the equipment. This includes: shipping, customs fees, warranty/service contracts, software licensing, on-site training. https://www.nserc-crsng.gc.ca/ResearchPortal-PortailDeRecherche/Instructions-Instructions/RTIInstruct-SOIInstruct_eng.asp https://www.nserc-crsng.gc.ca/ResearchPortal-PortailDeRecher... Note that travel, salary and benefits, consumables, renovations, lab infrastructure are not allowed. For other types of grants the institution's "tax rate" is usually specified by the Department of Research, and can be as high as 25%. UBC hides it's rates behind a portal, but Simon Fraser University has a description of them: https://www.sfu.ca/research/researcher-resources/fund-my-research/proposal-development/overhead.html https://www.sfu.ca/research/researcher-resources/fund-my-res... I won't speak to the amount of "administrative bloat" that universities have experienced during my two decades post graduation. While it exists, a tax rate of 25% seems quite reasonable when you consider the overall costs of research at a university compared to equivalent commercial office/lab space.

nkurz 1y ago
The numbers you are suggesting do indeed seem reasonable, and if those were the overheads being used, they would probably be justified. But I think current overheads are often much higher. Here for example is a recent article saying that the proposed 15% cap on overhead for NIH grants would be disastrous for universities like Harvard that currently have a 69% negotiated overhead rate: https://deliprao.substack.com/p/understanding-nihs-15-overhead-cap https://deliprao.substack.com/p/understanding-nihs-15-overhe... To be fair, I'm not sure if that author is using the same definition that you are. In the article, they clarify that a 69% rate means that a $1,000,000 grant has an extra $690,000 paid to Harvard. Still, the idea that the rates should be the same for all institutions and should be closer to the numbers you remember certainly seems worth considering.

nkurz 1y ago

The numbers you are suggesting do indeed seem reasonable, and if those were the overheads being used, they would probably be justified. But I think current overheads are often much higher. Here for example is a recent article saying that the proposed 15% cap on overhead for NIH grants would be disastrous for universities like Harvard that currently have a 69% negotiated overhead rate: https://deliprao.substack.com/p/understanding-nihs-15-overhead-cap https://deliprao.substack.com/p/understanding-nihs-15-overhe... To be fair, I'm not sure if that author is using the same definition that you are. In the article, they clarify that a 69% rate means that a $1,000,000 grant has an extra $690,000 paid to Harvard. Still, the idea that the rates should be the same for all institutions and should be closer to the numbers you remember certainly seems worth considering.