Using grants to make donations to institutions for using their resources/products?

Question

In my academic area I constantly use free software/products such as Linux, Python, SageMath, Octave, Wikipedia, etc.

I know that one can include software costs such as Matlab, Mathematica, Maple, Adobe stuff, Mac software, etc. But for free software/products I have not heard of anyone doing that. It feels wrong that if I use Matlab/Mathematica/Macs in my research, I can get public funding to pay MathWorks/wolfram/Apple thousands of dollars, but if I use Octave/SageMath/Linux they do not receive anything.

In other academic areas something similar can happen. I assume the same happens with "open" databases that operate with donations vs private databases that mostly operate as for-profit. Other examples could be Wikipedia, OEIS, LibreOffice, Gimp, etc.

I think that since a lot of research depends on that free software/product, they should receive funding (at least if you heavily use it). I know they receive funding elsewhere, but it is minimal compared to the money proprietary-software companies get (a lot from grants, by the way).

Do you know of cases where donations costs were included in grants? Do you think it is the ethical thing to do?

Answer 1

It is ethical to pay the price you are asked to pay; I'm not sure if it's ethical to get a funding agency to pay more than the advertised price, and your example could be understood as suggesting precisely that, even if the cause is good. I've never heard of an agency allocating funds for donations (imagine the potential for corruption!)

I personally feel there is nothing unethical about using a free product. Academics have many occasions to give back to the community without formal compensation: refereeing papers, editing a journal, reading theses, sitting on grant selection committee, organizing or participating in outreach activities etc. It's all part of the broader duties that come with job, and for those who care to keep score I found in the longer term that service roughly evens out: you contribute for free roughly the same you get back for free.

Answer 2

The grants I'm familiar with (mostly US NSF grants) are designed to support a specific set of proposed activities. I'm pretty confident that making a string-free donation of money from a grant to a free software/open source software (FLOSS) project would simply not be possible for these types of grants. The problem is not that it's wrong to support FLOSS projects but that it's wrong to give money directed for a specific task away in a string-free fashion.

With the grants I've applied for and received, there's an important distinction between what are described as direct and indirect costs. What you budget for in your grant are the direct costs of the research proposed. This includes tons of stuff like salary for people working on the grant, equipment used for the grant, supplies, etc. but these need to be justified in terms of the way that they directly support the research. Although they don't typically do it, the NSF can even take-back equipment after the work on the grant is over!

Of course, there are many indirect costs of doing grant funded research. Things like an office space, an office phone, heating of the building, Internet, etc, are costs that universities bear to support the research but that don't support the grant in any direct way. Indirect costs are typically covered through a fixed fee like a "Facilities and Administration" (F&A) fee. At my university, this is currently 55.5% of whatever my direct costs are. They are tacked onto the grant kind of like a tax and I don't ever see those funds.

Donating to support open source software that benefits all your projects and everything else seems like a classic example of an indirect cost. Unfortunately, it can be difficult, or even impossible, to budget for indirect costs beyond what a university charges as part of their F&A fee.

But this doesn't mean that you can't support FLOSS projects with your grants! I can see two approaches that you might try:

1. Have the university donate out of the indirect costs they collect from your grant. Work with your administration to convince them to support the projects you rely on. The indirect costs that your university collects are largely unrestricted. If you work with your university IT group or libraries—ideally before you even apply—you might be able to talk them into donating some portion of the indirect costs to the non-profits supporting your software. At my university, some discretionary portion of grants is returned to departments so you might even be able to convince your department chair to do so! This might be difficult but it's impossible. I know that many university libraries effectively donate to support the open-to-the-world arxiv.org.
2. Support the FLOSS projects you use directly as part of your research. If there are things that features that the software your using is missing or things it doesn't do that would be helpful to your grant, budget money to support the software developers as subcontractors for the grant. For example, you might budget for 3-4 months of Octave developer time. This way, the developers get money to support their projects, you get the stuff out of the tools that you need to help your research succeed, and you get to describe improvements on the software as outcomes or deliverables for the grant! Everybody wins!

Answer 3

I have not heard of anyone contributing grant money to an open source project. In the broad scheme of things, it's very ethical to donate money to support a product that you get value from. You're also helping to ensure that the product has a future, which is beneficial to your use as well.

I would not donate grant money to open source software without asking first, mainly because although it's a highly ethical thing to do, most people have become accustomed to using free software/resources without donating, and take it for granted a little bit. However, I would highly encourage you to ask whoever is giving you the grant if you could make reasonable donations to these foundations with the grant money. You could easily make a case to present not just based on the ethical value of support, but also on the research value of support—by donating money to open source foundations, you're supporting the future of a resource that is integral to the work you're doing. As an added plus, most foundations ask for much lower donation amounts than for-profit corporations charge for proprietary software.