Dollars for research and the NIH

Overhead is out of your reach

When compared to private enterprise, The current approach to NIH grant funding is incredibly inefficient at best which I will demonstrate with analogies. The participants in the process have had years to improve the efficiency of the systems, but they have not. One can only conclude that to them, the system does not need fixing.

On February 10, the NIH issued a directive limiting overhead expenditures on its grants to 15%. The institutional uproar was predictable and universal. For example, the University of Wisconsin-Madison declared, “This proposed change to NIH funding – UW–Madison’s largest source of federal support – will significantly disrupt vital research activity and delay lifesaving discoveries and cures related to cancer, Alzheimer’s disease, diabetes, and much more.”

Universities assert that every dollar that they are given is critical to prevent deaths in the streets, and that they are currently operating under the slimmest of margins. Does anyone outside a university believe this? It is likely overhead funds trickle into communications, public relations, hiring extra human resources personnel, dubious material and facility purchases, offsetting other costs like tuition or salaries, etc.

The current grant-funding process relies on the federal government and the universities to work together to keep costs down. Consider the cost of further education as an example of how these two entities have controlled costs elsewhere.

The existing system to determine overhead costs involves a negotiation between the research university and either the HHS or Department of Defense (DOD). The DOD is the organization that has been caught purchasing items like $100 screw drivers. It legally acquires 2-billion-dollar bombers. Cost-effective procurement has never been its strength. In this process, there is no proxy for the taxpayer, and neither party has a compelling incentive to reduce costs.

It is common for research institutes to trumpet successes, and there have been those, but research funding is as much a game of numbers as anything else. Most research does not result in radical gains. Science tends to produce incremental progress. A successful project moves science forward stepwise, and it is the hope that the sum of many small gains will produce a sweeping success. Whenever successes are called out, this selectively ignores all the dead ends and unsuccessful projects. When Nobel Prizes are given, they are usually given for decades-old research, which is only recognized in retrospect. Groundbreaking work may have been shunned or minimized in its time. To use a baseball analogy, successful research is as much a matter of increasing times at bat compared to home run hitting, and you can only determine home runs after the fact. Consequently, the best way to produce success is to increase the number of times at bat, and the way to do that is by increasing the ratio between the amount of money directly funding research to total dollars allocated, which requires reducing overhead.

Let’s compare grant funding to buying a car. Imagine, you’re buying a car, and you only have a fixed amount of money to spend. You have found the car you like. It’s well made, meets your needs, and is even in your favorite shade of blue. Your happy to find out it is listed at $10,000. You know that before you leave, the dealer will try to inflate the price. They will tell you the Extras are all essential-- dealer-applied rust proofing, an extended warranty, Scotch guarded seats, VIN etching into the windows, road-side assistance via helicopter, chrome-plated key fob, etc. You plan to haggle.

However, if instead of a car, this was an NIH grant, you do not have a choice about which add-ons are necessary. The dealership and some federal agency have already agreed on those choices long before you set foot in the door. Suddenly, your $10,000 car costs $15,500, and there’s nothing you can do about it. (UW-Madison’s mark-up is 55%). If you’re a consumer, you’d find another dealership and start over.

Now, what if you are a purchaser for a fleet of rental cars with a million-dollar budget? You need to buy as many cars as possible (to maximize your times at bat), and every $10,000 car turns into a $15,500 car. You can buy sixty-four of them. With a fixed 15% overhead, you could buy eighty-sixth that’s 22 more automobiles. More cars or more times at bat means more chances at discoveries.

I doubt that 15% overhead is the perfect number. History indicates that governments are terrible at controlling costs, and there seems to be no real representative of the taxpayer in the current process. We could reform the process to separate expenditures to build or maintain costly infrastructure into their own grants and use regional hubs to reduce the duplication of efforts. My hope is that the results of this step are a deep look at the process and real efforts to optimize the amount of grant money paying for the research, which is what the taxpayers and legislators intended. In an era of ballooning debt, even the universities may need to tighten their belts just a little.