People are the most expensive part of research
Breaking down what researchers cost in an academic lab
People are by far the largest regular expense in running a research laboratory. As a graduate assistant, or as a postdoc, I didn’t think much about where that money came from, or what it cost to fund my salary. The money I brought home in my paycheck was what I cost, right?
Wrong.
Before diving into the details of what researchers cost, there are some concepts that are critical to understanding how people are funded. Different institutions might have different language around this, but the core ideas are likely similar.
Full Time Equivalent (FTE): FTE is used to explain an employee’s percent effort on a funding source relative to a 40-hour workweek. For example, a full-time employee working 40 hours would be at 1.00 FTE (that is, 100% of a full-time 40-hour workweek). In contrast, a graduate student at 0.5 FTE equates to 20 hours of work (50% of a full-time 40-hour workweek).
Employee Related Expenses (ERE): This is the cost of an employee beyond their base salary, usually expressed as a fixed percent of their base salary. This is also known as fringe expense or fringe rate. Wonder why you get paid time off (sick time and vacation)? Or where your retirement or health benefits come from? They come from ERE. Importantly, the ERE percentage varies by position classification, reflecting the differences in the benefits associated with each classification (more below). These costs are typically paid from the same source that funds the employee’s salary (grants, start up funds, departments, etc).
Understanding ERE and FTE is important when constructing proposal budgets to fund the researchers conducting the research. It might seem like the order of operations in constructing a grant proposal is to determine what cool science you want to do first, identify what types of researchers are needed to do the research, then construct a budget around the science. While that’s partially true (you’ve got to have an idea to begin with), it’s really a waltz between the budget, the science, and the duration of funding. Often, you’ll have to adjust the scope of the project—or the types of personnel funded by the proposal—to fit funding limits or to make the project doable in the timeframe of funding offered.
Below, is a breakdown of what different types of personnel cost as they currently apply in my current department at the UArizona. While some of the expenses below are set at the university level (ERE, for example), others (like graduate student salaries) vary at the department or college level. These values (percentages, salaries, and other policies) often vary—sometimes greatly—across universities.
Principal Investigators (PIs; ERE rate 31.9%): Many professors that run research labs are on 9-month contracts. That is, they are paid for the 9-month academic year. They can obtain support for the three summer months as “supplemental compensation” from research grants at the same salary rate. For example, a professor that makes $80,000 on a 9-month contract, can bring in (up to) an additional $26,666 on grants should they be fortunate enough to have ample funding. The base $80,000 is paid by the University, the PI must solicit the remaining funds from grants.
In addition to the PI’s base supplemental compensation salary, the budget includes the ERE for their salary. The UArizona ERE rate for PIs is 31.9%. So, one month of an $80,000 salary is ~$8,889 ($80,000 salary ÷ 9 months). ERE is an additional $2,836 ($8,889 × 0.319 = $2,836). For a total of $11,725 per month ($8,889 + $2,836 = $11,725). So, it costs nearly $12,000 per month of supplemental compensation. If the PI’s base salary is higher, this cost is higher.
Many funders have upper limits on how much time (in terms of months), or dollars, can be spent on supplemental compensation for the PI. For example, the National Science Foundation limits this to two months of support across all NSF grants. Additionally, paying yourself above your rate is commonly not allowed. For example, in the above $80,000 example, you could not pay yourself more than $8,889 base salary per month.
Staff positions (ERE rate 31.9%). It’s often said that good technicians are hard to keep funded. This is because they are in the highest ERE class and a good tech should make a good salary. A double whammy. Using similar math as above, a technician that takes home $65,000 costs the grant $65,000 × 1.319 = $85,735. This is often why technical staff salaries are often pieced together across several grants. For example, a well-funded principal investigator may have three awards and split 33% of the technician’s effort (i.e. 0.33 FTE) across these three grants, such that each contributes equally to fund the researcher. I’ve heard that the ERE rate for similar positions at some universities can be as much as 80%, which makes it difficult to fund technical staff long term.
Postdocs (ERE rate 17.6%). Postdocs at UArizona do not pull the same ERE rate as technical staff because they are offered reduced UArizona benefits compared to other appointed professionals. (Something that I do not support). A postdoc making $65,000 take home pay costs a grant $65,000 × 1.18 = $76,700. That’s $9,035 less per year than a technical staff with the same take home pay.
Graduate trainees (ERE rate 13.0%). Graduate students are considered ‘trainees’ at most higher education institutions. As such, their benefits are typically greatly reduced compared to staff and postdoctoral researchers. This translates to a substantially lower ERE. Additionally, many institutions place bounds on allowable graduate student salaries and FTE. During the academic year, UArizona graduate students are capped at 0.5-0.66 FTE, depending on their citizenship. During breaks, the limit increases to 0.875 FTE, although individual PIs may choose to pay students at a lower rate. In my department, a doctoral graduate assistant is paid from a 9-month 1.0 FTE base salary of $45,964. But during the academic year, they are capped at 0.5 FTE, so their take home salary for the academic year is $45,964 × 0.5 = $22,982. If they work in the summer at 0.875 FTE (406 hours), that is an additional $13,486, for a total annual take-home salary of $22,982 + $13,486 = $36,468. Then we add 13.0% ERE on top of this: $36,468 × 1.13 = $41,208. The cost of graduate students doesn’t end here, as we also pay their tuition on top of their base salary and ERE. At UArizona, the graduate student tuition for is typically paid from a grant. The current annual tuition rate for ≥0.5 FTE students is $12,718. For every year of salary ($41,208), we need to add $12,718. That totals to $53,926 per student, per year. Tuition is paid even after their formal coursework is finished and the student is taking research or thesis credits. Some universities waive tuition for graduate students, which makes funding their training a more economical.
Undergrads (ERE 2.0%). Undergraduates are typically hourly employees and capped at an upper FTE limit. At UArizona these limits are 0.5 FTE during the academic year and 0.875 during the summer. Their fringe rate is low because their positions as student workers are not associated with fringe benefits. We start undergrads at minimum wage and increase their wages as they grow in their skill sets, usually on an annual basis.
Below is an example of the personnel costs for a proposal from a few years ago. The total proposal value was ~$675,000 and the personnel cost alone was just under $410,000. That is, funding the person-power was 60% of the total budget.
As we climb through our academic careers, our perception of the cost of science changes. As a graduate student, I envisioned a $100 reagent as a crippling cost for the lab. Looking back, I probably thought this because $100 would have been a massive cost to me in my personal life at the time. In reality, that $100 is minuscule, relative to the cost of the person doing the research with that reagent.
An example of this thought process at work can be seen looking at the cost of high-throughput DNA extractions in microtiter plates vs classic DNA extractions in single tubes. Single-tube DNA extraction kits are ~$0.21 per reaction cheaper than the microtiter plate-based extractions. But when you factor in the amount of time saved across 96 samples viewed through the lens of the cost of someone’s time spent doing the extractions, the cost per reaction is much reduced—even if undergraduates are conducting the extractions—simply because it takes less time to run the extractions in the high-throughput format. (Of course, this example is only a meaningful savings if you have many DNA extractions to conduct).
As we climb through the academic ranks, we’re taught (both directly and indirectly) that our time is not valuable. But it is. I always ask “how much time will it save you?” when someone asks for a new tool, no matter the cost. And I try to weigh the utility of that tool over time—in the context of personnel hours it saves—when deciding whether to make the purchase.
So often we forget to factor in the most expensive cost of running a lab—people— when making decisions about what tools to buy or not.
Thanks for writing this piece. I am a master's student in India with dreams of being a PI one day. This really was very relatable in the sense that I too considered the machines and the reagents to be the major cost drivers.
like rest of your posts this was helpful with a fresh new better perspective. So much tacit knowledge is their in academia it needs to be shared and spread, thanks for writing.