# All funding organizations are research organizations [[Research management matters]]. Many people hold an implicit idea that research management and research funding are decoupled. The story goes that researchers (whether organizations or individuals) propose projects that they will then manage and funders decide who to fund. This is wrong. Funding organizations are managing research whether anybody wants to acknowledge it or not. Any filter on what gets funded is a type of management: in addition to the obvious fact that work that doesn’t get funded will happen less frequently, researchers will propose things that they think will get funded, which in turn shapes what kind of research ideas even get thought.[^2] [[The way people think drastically changes depending on their context]]. Until [[Focused Research Organization]]s became “a thing” nobody would have even proposed something like [[The Rejuvinome Project]] to a serious funder. Reporting requirements are another way that funding organizations are in the management business: research takes very different trajectories if it needs to show progress every quarter. At the same time, there is valuable work that doesn’t happen because funding organizations are reneging on their management role and researchers bristle at the idea of being managed. Funding is a critical coordination mechanism for work that requires coordination between multiple organizations. Without some hierarchy, individual organizations (whether they’re labs, companies, or just individual researchers) only coordinate when doing so is heavily aligned with the games they are already playing. ([[People and organizations are all playing some game that has different ways of gaining status and power]]) Labs will coordinate when it leads to a prestigious paper, companies will coordinate when it leads to revenue, etc. [[Funding can be seen as buying time from some game to enable someone to play another game]] — enabling a professor to buy that new piece of equipment in exchange for work that is less publishable, for example. Funding is a unique coordination mechanism because of its ability to shift games so directly. Everybody has their price. Funding is a powerful coordination mechanism. A large part of DARPA’s modern success happened because they pushed researchers to work on things they would not otherwise have worked on in conjunction with other researchers across several disciplines and industries (see: autonomous cars, advances in electronics, etc.) In the past, other government agencies (like the Energy Research and Development Administration before it was disbanded in the early 80s[^1]) played a more opinionated role and acted as a hub between different research groups — iteratively figuring out what was feasible and how to nudge research projects towards a higher-level goal. Before the era of state-dominated research, systems-level coordination was less important, but wealthy individuals like [[Alfred Loomis]] still brought together researchers in places like [[Tuxedo Park]] and nudged them to collaborate on things he thought were interesting, resulting in fundamental work on radar and other systems. Research coordination can certainly happen without money involved, but it is much weaker and harder to do, especially in the hyper-competitive environment we find ourselves in today. Organizations that fund research are research organizations regardless of whether they directly hire the hands doing the experiments. In situations that need heavy coordination, organizations need to both do research design and actively manage different projects. Calling DARPA a “funding organization” is a misnomer. In many cases DARPA PMs do not simply put out a call for applications and pick which ones they like. Instead, they first create a concrete vision for what needs to be done, work closely with researchers on project ideas, and then make sure those projects coordinate towards a broader goal in a way that would never happen if the researchers doing the projects were left to their own devices. Of course, even within DARPA, the involvement in the process sits on a spectrum. But management is a form of coordination, and all funding organizations manage research whether they admit it or not. That management makes them part of the research process. Thinking of research management and research funding as decoupled is not only wrong, but harmful. It incorrectly absolves funders of responsibility for being good managers and the critical role they play in activities that require cross-organizational coordination. Conversely, it gives many researchers expectations of autonomy that sabotage higher-level research goals. I suspect [[Donald Braben]] would say that this notion of funding=management=research is bullshit. [[brabenScientificFreedomElixir2008]] makes the case that funders shouldn’t go near management with a ten-foot pole — even reporting requirements and judging the feasibility of proposals hamstrings the process of scientific exploration. Many researchers would agree — one of the most striking results in [[collisonWhatWeLearned2021]] was the fact that `In our survey of the scientists who received Fast Grants, 78% said that they would change their research program “a lot” if their existing funding could be spent in an unconstrained fashion. `. However, Braben’s position isn’t actually mutually exclusive with funders as research managers. As any good manager knows,[^3] different people and projects need different management styles. There are incredibly important people and classes of work that are crushed by any constraints. Imagine Einstein’s grant proposal and progress reports for the eight years it took him to formulate [[General Relativity]]. At the same time, there are people and goals that get nowhere if everybody pursues their own agenda. It’s a big, messy spectrum with ample room for argument about how many and which constraints are important. Which activities benefit from which kind of management is a deeply understudied area of metascience. My hunch is that a more rigorous treatment would find that research targeted at building radically new systems needs more direct management to coordinate different components and make sure they’re integrated, while research into core mechanisms of biology and chemistry needs *less* management to allow researchers to quickly pivot to new avenues of exploration that don’t seem promising to an outsider. Any approach is a management choice, regardless of what that choice is. It’s misleading to draw a bright line between organizations that fund research and execute on research. Research management matters, and any organization participating in the research process, even if it’s doing nothing but giving away money, is doing some form of research management. More radically, the idea of “funding organizations” is misleading, reenforcing the proposer/funder dichotomy and collapsing the space of possible management styles. Acknowledging the inevitable importance of research management could enable people within organizations to mindfully shape their management to enable the specific activities they want to see in the world. ### Snippets Just as there’s a spectrum of management appr There is a spectrum between giving an individual or organization funding to do whatever they want to hiring a contractor to hiring an employee (and employees have a similar spectrum). It’s also true that there are many ways around the system on both the funding and performing front. Different organizations will (and should!) focus on different activities There is indeed a distinction between organizations that hire people who are executing on a research agenda and those that do not, like DARPA, the NSF, or Fast Grants. But the mistaken If research management does indeed matter, it means that the common perception that “research enablement = research funding” is wrong. To focus only on funding is to implicitly say that filtering-based research enablement is the *only* option. There’s an implicit attitude that coordination should always happen at the the best way to enable good research is simply by finding high quality people/projects and getting them money to do their own thing. It’s important to acknowledge is The thing that makes DARPA work is *good research management*. The same thing could be said about Bell Labs. The balance of freedom and focus The missing management on the part of funders has also led to a rent-seeking system. A repeated interaction I see is that professors have their own agenda that they want to sell — the game is whether they can convince a funder to finance that program. *Rent seeking* Giving out money *is* research management. It might be possible to think of research management as having two axes This is tightly related to the [[Heilmeier — Licklider — Braben — Elon quadrants of Research Management]] If you’re only giving out money you’re only able to actuate along the enablement axis. Some people might object to the term “research management” — that scientists should be left to their own devices and not managed. I agree! But no intervention is in fact a management style. ### Related * [[Program design could be done as a residence program]] * [[Activity design indicates how power and coordination are distributed]] * [[Vision is not enough]] * [[Grants only go so far]] * [[Activity design indicates how power and coordination are distributed]] * [[Motivation is not context-free]] * [[An important question is who is doing the research design]] ### References * [[odlyzkoDeclineUnfetteredResearch1995]] * [[A new structure for scalable research]] * [[barberAdvancedResearchProjects1975]] [^2]: There have been proposals for randomized funding, which certainly does reduce the amount of filtering, but unless you want to give the [[Timecube]] guy an equal chance of funding as anyone else, there will still be some filter. [^3]: I am not (yet) a good manager, but I don’t think this statement is very controversial. [^1]: See [[Stephen Dean podcast 20 Aug 2021]] and [[deanFusionPowerMagnetic1976]]