Animals in Research: Unnecessary Duplication

The Speaking of Research group has posted an open letter criticizing an analysis of research redundancy which forms the backbone of a USDA complaint authored by an ARA extremist*. The open letter points up the essential vapidity of the complaint and I encourage you to go read. (There are links to activist sites there so you may wish to be cautious about either providing it any more traffic or providing your IP address.) Returning to a few prior posts on the responsible and well regulated use of Animals in Research, I note that I never took up this most interesting topic for discussion, namely the nature of “unnecessary duplication” of research using animals.

Allocating Costs To Federal Research Grant Awards

Female Science Professor has two interesting posts up yesterday and today concerning allocating investigator effort and other costs to Federal research grants. Her basic gripe is that university and granting agency officials are forcing her to lie and/or pretend that she can accurately measure actual percent effort on a specific Federally funded project and can promise that software licenses (or equipment) charged 100% to a particular award can never be used for any purpose other than the direct furtherance of the specific aims of that award.
In relation to allocation of effort, she relates a discussion with a university grant management official about her percent effort on a particular project as follows:

And so on. We did not reach a number. I pleaded with her to tell me what number would seem like a good number and she refused because we might be audited and it would be bad if I worked more than what was listed and it would be bad if I worked less than what was listed. Every time I suggested a number, she rejected it because it was either somehow not an acceptable number (too low, too high) for mysterious accounting reasons or she wasn’t convinced it was accurate. I said that if she told me a good number, I would promise to work that amount, although I was of course lying because I’m going to work whatever amount is best for the project in the time I have available, and also I still don’t get the concept of time and effort being different but the same. She refused.

In relation to allocation of the cost of a specific software license to be used to analyze experimental data, she complains:

As regular FSP readers surely know by now, some of these rules and regulations make me crazy.
Example: I found myself promising recently that I would never ever ever use a software site license that I purchased with one grant for any activity not related to that specific grant.

The university and/or governmental officials who are putting FSP in the position of thinking she needs to lie and/or pretend about allocation of effort and other costs to Federally funded research grant awards are ignorant of the Government’s interpretation of the relevant statutes and regulations that governs allocation of costs. The relevant guidance is provided by Office of Management and Budget Circular A-21, which provides in pertinent part:

If a cost benefits two or more projects or activities in proportions that cannot be determined because of the interrelationship of the work involved, then, notwithstanding subsection b, the costs may be allocated or transferred to benefited projects on any reasonable basis, consistent with subsections d.(1) and (2).

This provision explicitly treats the situation where investigator effort or other cost benefits more than one project, but where the relative proportion of the benefit of the cost to the multiple projects cannot be explicitly determined. In this situation, the effort or cost may be allocated on any reasonable basis.
In the case of effort allocation, it would thus be perfectly acceptable for a PI to allocate her effort equally among her multiple grant awards, or to allocate her effort in proportion to the size of each of those grant awards, or in proportion to the number of other personnel being suppported by each of those grant awards. It would even be reasonable for the PI to simply make an estimate on no explicit basis of the effort devoted to each grant.
As a practical matter, the only way a PI’s allocation of effort is going to cause a problem in an audit is when the allocation is absurd on its face. For example, obviously you cannot devote more than 100% effort total to all of your professional activities. And if you are a teaching faculty, you also cannot allocate 100% of your effort to sponsored research.
More subtle cases would be like a Provost allocating 90% of her effort to sponsored grants. It is is simply not feasible that it only takes 10% effort to be a Provost.
In the case of other cost allocation, if you purchased a software license (or piece of equipment) with the purpose at the time of purchase to use that license for pursuing the aims of the grant that the license was charged to, and that you later figure out that the software can be used for some other purpose in your lab related to other projects or activities, that is absolutely fine, so long as the allocation of 100% of the cost of the license to that grant was “reasonable” at the time it was charged. If you purchased the license to perform necessary analysis of data relating to the charged grant, and you then performed all of that analysis, then it was totally reasonable to charge 100% of that license to that grant, regardless of what you do with the software above and beyond that analysis.
University or Federal agency officials who demand an explicit quantitative basis for allocation of effort and/or costs in situations where the interrelated nature of scientific research projects in a laboratory make such an explicit computation impossible are failing to follow the guidance of OMB Circular A-21.