Why does my academic lab keep growing?

Andrew, Breck, and I are struggling with the Stan group funding at Columbia just like most small groups in academia. The short story is that to apply for enough grants to give us a decent chance of making payroll in the following year, we have to apply for so many that our expected amount of funding goes up. So our group keeps growing, putting even more pressure on us in the future to write more grants to make payroll. It’s a better kind of problem to have than firing people, but the snowball effect means a lot of work beyond what we’d like to be doing.

Why does my academic lab keep growing?

Here’s a simple analysis. For the sake of argument, let’s say your lab has a $1.5M annual budget. And to keep things simple, let’s suppose all grants are $0.5M. So you need three per year to keep the lab afloat. Let’s say you have a well-oiled grant machine with a 40% success rate on applications.

Now what happens if you apply for 8 grants? There’s roughly a 30% chance you get fewer than the 3 grants you need, a 30% chance you get exactly the 3 grants you need, and a 40% chance you get more grants than you need.

If you’re like us, a 30% chance of not making payroll is more than you’d like, so let’s say you apply for 10 grants. Now there’s only a 20% chance you won’t make payroll (still not great odds!), a 20% chance you get exactly 3 grants, and a whopping 60% chance you wind up with 4 or more grants.

The more conservative you are about making payroll, the bigger this problem is.

Wait and See?

It’s not quite as bad as that analysis leads one to believe, because once a lab’s rolling, it’s usually working in two-year chunks, not one-year chunks. But that takes a while to build up that critical mass.

It would be great if you could apply and wait and see before applying again, but it’s not so easy. Most government grants have fixed deadlines, typically once or at most twice per year. The ones like NIH that have two submission periods/year have a tendency to no fund first applications. So if you don’t apply in a cycle, it’s usually at least another year before you can apply again. Sometimes special one-time-only opportunities with partners or funding agencies come up. We also run into problems like government shutdowns—I still have two NSF grants under review that have been backed up forever (we’ve submitted and heard back on other grants from NSF in the meantime).

The situation with Stan at Columbia

We’ve received enough grants to keep us going. But we have a bunch more in process, some of which we’re cautiously optimistic about. And we’ve already received about half a grant more than we anticipated, so we’re going to have to hire even if we don’t get the ones in process.

So if you know any postdocs or others who might want to work on the Stan language in OCaml and C++, let me know ([email protected]). A more formal job ad will be out out soon.

36 thoughts on “Why does my academic lab keep growing?

  1. The stochasticity of funding, which leads to the problems you describe, is discussed much less than the amount of funding that’s out there, though both are (in my opinion) similarly important. It’s very hard to run a group of single-digit-number size and have the fluctuations of money and people line up. It’s easy to run out of money, to have money and be forced to grow (as you write) though growing may not be pleasant or really productive, or to have a surplus of money and a shortage of people (my problem at the moment, which I can’t really complain about but which is harder to fix than I would have thought).

    • A lot of this comes down to the inability to save money for future use. If you get money for a 2 year period, you must have spent it by 2 years, or maybe ask for a 3rd year extension in some kinds of grants, but basically there’s no way to say get an extra grant and decide that right now growing your group isn’t a productive way to use that grant so you’ll save that grant funding for next year when your postdoc will be leaving anyway and it will make sense to hire a replacement to work on that project etc.

      There are lots of things wrong with funding today… Some of them could be improved by *adding* stochasticity (mainly in the area of deciding which grants get funded and hence avoiding the “hot topic” biases) but there’s plenty to be said for reducing stochasticity in individual lab cash flows… For example offering a kind of Basic Income specific to research scientists, so that not all funding is based on grants.

      • “A lot of this comes down to the inability to save money for future use.” I agree! I’ve always found this odd, since requiring that money be spent and not saved for later use incentivizes waste. Lots of equipment gets bought, for example, not because it’s needed but because money exists now that can’t be kept for later.

        • I’ve always found this odd, since requiring that money be spent and not saved for later use incentivizes waste.

          This is just how government spending works in general. If you don’t blow the budget you will get less funding next quarter/year/whatever. The odd part to me is that people concerned about the consequences of waste (eg climate change) want to give the government *even more money* to be used in this way.

          Things could be worse though. I was recently reading a book about Nazi Germany in the 1930s that said people would buy brand new trucks, remove the tires, and then sell the trucks as scrap metal. Imagine millions of trucks being manufactured and destroyed without ever being used for anything because government incentives made that the cheapest way to source rubber.

        • “This is just how government spending works in general. If you don’t blow the budget you will get less funding next quarter/year/whatever.”

          Ah yes, it always amazes me how little people understand about the Federal Government and how the budget works. The legal decisions on the budget are not that you MAY spend the amount budgeted, but that you SHALL spend the money budgeted and on the items they were budgeted for (Color of money laws). Not spending the money is considered going against the stated will of Congress. Our lab’s budget is monitored I believe monthly for both over-spending and under-spending and if you are in variance of more than a certain percent of your budget plan then you have to justify why, and yes that applies if you are under-spent.

          If you want to see spending changed on certain items, talk to your Congressional representative or Senator. Federal employees are legally obligated to spend the money allocated by Congress.

        • This is also how corporate spending works in my experience. The reason is that people give you money for some end, which is never to grow your bank balance. Companies want to see their capital put to good use by managers or they’ll find new managers that will spend their budget. I had a terrible department head once who thought he could please his managers by going under budget. That just means a smaller budget next year.

        • Sure, growing your bank balance isn’t why people pay you, but at the same time, it doesn’t make sense to require that bank balance to drop to zero at the end of a certain period, particularly if there’s not any kind of guarantee of renewal of the bank balance. The concept of “float” is an important part of good financial management. Requiring people to spend every last cent by a certain date is basically a requirement for financial ruin. Managing a healthy float balance is something we should be encouraging, sure having the float grow without bound would be bad, but having a balance that tides you over between grants without firing everyone in the lab and spending 100% of your time writing grants until you are re-funded is obviously far more efficient and accomplishes a lot more of “some end” than otherwise.

        • It might be a good idea to allow flexibility in the duration of the grants: instead of a one-year grant, it’s one to two years; instead of a two-year grant, it’s two to three years. If you apply for a bunch of two-to-three year grants and get more of them than you expected, you stretch them all to three years and don’t start applying for new grants until a year from now. If you get fewer than you expected, you start applying for new grants right away and in the mean time you can spend your grants out in two years. In practice all of your grants would have staggered start dates and ranges of end dates, so this is the functional equivalent of being allowed to have a modest floating balance, with the advantage that there doesn’t have to be a central authority checking your balance since it’s an automatic outcome of having grants of flexible duration.

        • Yes, this could help, but I also think it might be wise to explicitly provide some amount of “float funds” along with each grant. So for example you apply for a $500k grant over 2 years. The government already adds say 50% on top of this to give to the institution as “overhead”. In addition to that it would make sense to me to allocate say 10% to “float” which basically goes into a PI’s float account and has no spending deadline at all. You could for example cap a float account at say $300k (inflation adjusted) with ease.

          The assumption should be that anyone running a lab hiring multiple people and doing multiple projects should be able to save enough to cover temporary shortfalls of 3 to 6 months salary for example.

        • It’s not that uncommon for departments or universities to provide “bridging funds” already, but it puts a huge amount of power into the administration’s hands, to give out such funds to “people they like” and withhold them from “people they don’t like” or whatever. To me it makes more sense to put this control into the hands of people who are writing the grants that get funded (the PIs) rather than University administration.

        • Also my understanding from my wife is that NSF has moved to a rolling grant admissions schedule. You just apply for grants whenever you want, and when they get enough applications to make it worthwhile to send them off for review, they do it.

          Supposedly this actually reduced the application rate, and increased the quality, because now instead of jumping on an idea to meet the deadline and ensure they aren’t caught out with their financial pants down, people can apply at whatever time makes sense from both the quality of the grant perspective as well as when they actually will wind up needing the money.

        • For NSF and NIH, it’s already quite standard to be able to get a one-year no-cost extension. This is good, but one year isn’t really that significant for the issue of stochastic “starts” brought up in the post.

          About rolling / no deadlines for NSF: this doesn’t make it easier to get money when one needs it; the review panels are formed when there are enough proposals to warrant review (roughly speaking), and *then* the “usual” 6-9 month process begins. It is certainly fascinating that submissions have gone down with this process in place, but I think that’s telling us something about human responses to deadlines rather than budgeting plans!

          About allotting 10% of a grant to a PI’s “float” account — this would be wonderful, and would make many things run more smoothly. When I’ve brought this up in the context of overhead, the response is that such a use of money would be explicitly prohibited. I think it would have to be done by the funder.

        • Raghuveer: the NSF rolling submissions hasn’t been going on long enough to really figure out how it will affect budgeting, but I do expect it to improve things over the long run. I agree with you about 1 year extensions not being enough to handle the stochastic start time issue.

          The float account idea would have to come from the funding agency, but I think it would be a smart thing for them to do to improve efficiency all around. Every hour spent applying for a grant you don’t get is an hour you’re definitely not researching anything much… reducing the urgency of the stochastic start and potential zero balance end would improve research quality and efficiency.

  2. I would think that your lab would have a natural ceiling on how many projects could be undertaken due to the limited time of qualified PI’s in the lab. In other words, if you, Andrew and Breck are regularly bid as PI’s, and if all of you also have teaching duties or other such commitments, then writing too many proposals would mean bidding one or more of you for more than 100% of his time.

      • Your problem is not instability, it’s excessive stability. Nature provides a model for keeping growth in check through disruptive forces, and you clearly have built a system that is (for now) highly resistant to those forces. The analogy is between a lab of researchers and a pack of wolves or a band of gorillas or a pride of lions, etc. There are at least two mechanisms that regulate the size of these groups: competition for scarce resources and the imperative to disperse DNA as widely as possible in case the group gets wiped out. To complete the analogy, you are growing without restriction because you don’t have (good) enough competition for grant money, and because you are not spinning off other labs. I know you’re putting well-trained and like-minded post-docs out into the world, but it takes a long time for them to get advanced enough in their careers to have labs that are competitive on your level–federal funding is generally a zero-sum game, especially given the rich-get-richer effects of promoting synergy.

        To switch analogies midstream (and mix a metaphor in there to boot!), you could do with some intentional, well-planned pruning, so that your seedlings can grow. In the most extreme version of this, one of the three of you leaves to set up an independent shop. A more moderate approach would be to actively work to spin projects off into other labs led by former post-docs, so that mature but less central lines of research end up with PI’s at other institutions. In the short-term, this could be painful, and could even lead to your lab not getting enough resources. But in the long term, the philosophy/methods/ideals/goals of your lab, its DNA, will survive your lab’s inevitable (sad but true) demise. If you stay the course, then to quote Jeff Goldblum, “Nature finds a way.” If you do not intentionally introduce instability, it will happen unintentionally!

  3. What are the limits to how many grants you can apply for? Could you apply for 100 grants, and get, maybe 40? Are the limitations that a given agency is not going to give you 100 grants, and that there are not 100 grant-giving agencies? Is there coordination among grant-givers to prevent this problem? Do you have to disclose your other funding when you write a grant to prevent this?

    Do some academics basically run grant-writing factories since the cost to writing a grant is not very high?

    Do you have some sort of equity stake in the lab, or is your payoff in prestige (the coin of the academic realm).

    • Terry:

      There may be labs in biomedicine that have 40 grants. I guess it depends on what you call a “lab.”

      We don’t directly make money from our lab, except for salary, but I guess we make money indirectly in that, if our research is successful, we are invited to consult and give paid lectures.

      The reason I apply for grants is not for the money I make from it, and not for the prestige, but because there are things I want to figure out, and work I want to get done, and funding can pay for people to do this work.

    • Terry – in answer to some of your questions. It would be very hard for an individual to apply for more than 10 grants per year. In the health area, there are a handful of large grant-giving agencies (NIH, NSF, FDA, PCORI) in addition to some large charities. At my institute, the bulk of the money comes from large, long term U grants from the NIH. These run 4 years and can be renewed. I am not sure if Andrew has access to those type of grants which can be renewed. The large agencies do require disclosure of all other grants.

      I guess you could call our institute a grant-writing factory since we are 100% funded by grants. Faculty in our institute are ‘encouraged’ to get additional funding so it is always on one’s mind. It is a tough business and except for running very lean in terms of staff, not sure how one can get more stability.

      • Yes, we could apply for 100 grants as a group. We apply for around ten per year of different sizes or more if you count grants where we aren’t the primary institution. The limiting factor is staff to do the work. Proposals require you to justify how you’ll do the work. Also, we don’t want to grow!

  4. What are you proposing to do in these various grants? Are you ‘double applying’ on the same work? If not, if you’re proposing new/different work on each grant application, I don’t see the problem. If you are, maybe you should stop (?).

    • Alex:

      For these grants we’re proposing to do all sorts of cool and important things. The problem is only that if we don’t get enough funding, we can’t pay everybody, and if we get lots and lots of funding, we have to hire new people. Planning is a challenge. This is not at all a problem unique to academia: it happens in just about every business. Indeed, I suspect the only reason we find this situation such a challenge is that so much of academia is stable: roughly constant numbers of students from year to year, roughly constant teaching loads, an approximately fixed number of professors in each department, and so on.

    • It’s called “synergy” and funders like it. It demonstrates you have the resources to get things done. We can’t apply for exactly the same thing multiple times, but almost all the work we do is complementary to our other work.

      The problem is how to keep our group funded at the current size.

      • Suppose for the moment that Universal Basic Income was a thing… everyone in the country got some basic monthly income and certain people got additional amounts to pay for specific issues, like for example people with chronic disabling health conditions, or people who provide medical services to rural areas, or people who participate in natural disaster planning and response programs, or whatever.

        Now, suppose one of the categories enabling you to get a higher basic income was to hold a job involving certain types of research. In essence, if you are creating a public good like freely disseminated knowledge, in addition to whatever your salary was you qualify for something like 50% of GDP/capita per year additional tax free, that’d be somewhere around $3000/mo currently.

        Grant money now would be only paying the marginal increase required to retain people in this type of research. Suppose that take-home after tax of say $6k/yr is typical for researchers. If they’re taking home $3k in UBI increment, plus say $1k in core UBI, we’re talking about only needing to pay on the order of $2k after tax in salary, taxes under this kind of scheme would need to be around 30%, so let’s call it $3k/mo from the grant, that’d be $36k/yr compared to currently say $120k/yr which could be split between paying the researcher UBI increment and funding more grants.

        In any case, these are just back of the envelope numbers to make the general point which is that for people willing to produce public goods, guaranteeing a moderately comfortable basic level of income for them could pay dividends by getting them to work more on the projects of interest, waste less time on writing grants, and ensure stability of their productivity even between periods where grant funds were temporarily scarce.

        If the basic notion that the government should produce public goods like knowledge makes sense, there isn’t any law of nature that says that the best way to fund such things is the granting process we have now. In fact, there are good reasons to think that the granting process we have now favors say the kind of bogus research that Wansink and others did. Obviously abusing any system is possible, so thinking hard about how to implement a UBI for researchers would be a good idea, but the notion has appeal as it would seem to inherently reduce the clique power of grant concentration.

        • sorry, when I say “take-home after tax of say $6k/yr” I meant $6k/mo which is $72k/yr after tax, given FICA, federal tax, and state taxes, that’d be something like $116k/yr pre-tax which is probably similar to real researcher salaries.

        • creating a public good like freely disseminated knowledge,

          Define “knowledge”. To the vast majority of researchers today “knowledge” is “a statistically significant relationship exists (or does not) between two variables”. And those researchers are considered the experts by the other 99+% of the population. We do not want to produce any more of this intellectual pollution masquerading as “knowledge”.

          I am sympathetic to the UBI idea*, but why not just do it without any strings attached? The beauty of a UBI is supposed to be conserving resources by cutting out a bunch of unnecessary bureaucracy.

          * Actually I managed to make a bunch of “fuck you money” in the middle of my graduate training, which gave me a lot more freedom to do what I wanted. This essentially functioned as a personal UBI. It is a whole different world when you want to do research but don’t care so much about getting paid for it. My impression is this circumstance was another thing very common in pre-WWII academic research (mostly conducted by the upper class).

        • Some level of UBI should absolutely be no strings attached…

          I also agree with you about the difficulty of dealing with gaming the system, and the kind of “knowledge” Wansink created was absolutely a mechanism for gaming the system… so the problem exists with or without the UBI idea.

          There is plenty to discuss on how UBI should be administered… but in this more restricted context, my main point was that it might improve efficiency if we give out some of the funds that support research unconditionally to people who somehow through some level of screening show that they are working on producing public goods. The whole concept of giving money to research is to fund public goods, but why do 100% of it by competitive grants? Is that the most efficient way? There is a huge bureaucracy associated with administering grants. Would such a thing be as huge to simply verify that people had say held some kind of legitimate research position for the previous 2 years in order to qualify to receive some but not all the necessary funding for their research?

        • I’d look into how things were done pre-WWII. Wasn’t it mostly universities funding the research via endowments?

Leave a Reply

Your email address will not be published. Required fields are marked *