After a decade of running my lab, I decided to crunch some data one night after the kids were in bed. No, silly, not scientific data. I wanted to answer a nagging question: just how poorly, in quantitative terms, had my lab been treated by collaborators?
As a computational team, we collaborate with biologists on every project. Researchers often seek our expertise in image analysis, data science and deep learning, and ask me to contribute to their grant proposals – often just a few days before the deadline. I advise on experimental design and contribute some text, data and a supporting letter. The collaborator requests funding for my lab’s future work on the project, typically about $50,000 (£38,000) out of a budget of more than $1 million.
But sometimes that is the last that I hear – and not because the proposal was rejected. The conversation often goes like this:
Me: “Hey, was your proposal funded? Yes? Cool, let’s get started!”
Them: <Silence for 4 years>.
So, to the data crunching. Over the past 11 years, I contributed to 28 collaborators’ grants that were ultimately funded (never mind all those that weren’t). One-third of those funded scientists provided zero funding for my group, and another third cut the budget, providing only 10 per cent of the proposed amount on average.
In fairness, we were not (usually) asked to perform the proposed work in these cases. And the main awardee is ultimately responsible for success, so they must be free to evolve their scientific strategy. Most funders permit this, particularly if the funder cuts the overall budget. With cuts often topping 15 per cent, the awardee must make tough decisions, and often collaborators are the first to go. Believe me, I understand how challenging it is to cope with budget cuts!
But none of these realities excuse failing to communicate a decision to drop collaborators. It’s one thing to say, “Sorry, we are changing direction on the project – we don’t need your help any more.” It’s another to just take the money and run, whether intentionally or passively. After all, collaborators are scientists, too, with their own research and finances to juggle.
When I mentioned the problem of collaborator-ghosting online recently, it became clear that my experience is far from unique. No doubt the frequency of bad behaviour varies depending on the type of collaboration. Awardees often see computational or data effort as a luxury, and might decide to struggle on their own – probably resulting in inferior analyses. I also suspect that scientists are more likely to keep commitments to collaborators within the same institution. And the demographics of the people involved probably also matter a great deal: those imbued with more power face fewer consequences for bad behaviour, and less backlash when asserting their position.
If we want interdisciplinary science to thrive, researchers must stop treating collaborators as disposable. Investigators may not realise that they are doing this. Maybe they don’t appreciate the impact that it has on the labs that they leave hanging. But if you change scientific direction and shift budgets, at least tell the collaborator. And don’t ever include a collaborator for name recognition or expertise in a fancy technique when you intend to cut them later (known as the “trophy collaborator”, “bait-and-switch”, “push off the boat”, or just plain “jerk” approach).
For those who find themselves on the receiving end of this nonsense, what to do? Colleagues have suggested various remedies. One is to demand that collaborators keep their commitments. This is easier said than done, particularly if there are power differentials involved. Besides, if you need to stamp your feet just to begin working together, it does not bode well for the collaboration. It is also unclear when to begin pushing: the timeline of the project is rarely set in stone, so the awardee can easily say “not yet”, until the budget is fully spent.
Another suggestion is sometimes effective: using funding agency mechanisms to enforce commitments. For example, it is difficult to change budgets on a National Institutes of Health “multi-principal-investigator” grant or remove a collaborating PI named as “key personnel” – although the awardee can entirely cut funding for the PI’s supplies and staff. Funding agencies should consider evolving their mechanisms and budget-cutting habits to suit modern science, which relies so heavily on teams.
Perhaps the best advice is simply not to count on collaborations to fund substantial amounts of your own laboratory’s work, even if it is all collaborative research. But I also plan to adopt a fourth tip: whenever people ask me to contribute to their grant proposals, I will set expectations for budget changes – or maybe I will just send a link to this article.
I hope that these remedies work. My team and our research thrive on collaboration. And, more broadly, it would be a real shame if budget-cutting pushes academics away from working together. We know that, particularly across disciplines, it can produce the very best science.