DOGE and Federal Budget Cuts

I’ve been mulling for some time about writing this column, but have avoided it. What finally pushed me over the edge was an email from NSF titled “Focus on Recruiting Emerging Climate and Adaptation Scientists and Transformers”. That I’m annoyed has nothing to do with “climate” in the title, even though, as readers here might suspect from a previous guest column, that I’m more in the Lomborgian school of how to deal with climate change (deal with it and recognize that it’s not all bad) than the Thunbergian school (run around in a panic like Chicken Little). Instead, it’s the focus on anything but science that galls me. What is “adaptation science”, anyway? Your house burns in a wildfire because the state didn’t keep up fire-prevention efforts, you couldn’t get insurance because of overregulation, you deal with it. No science involved (OK, I’m being a little snarky, so I guess I can imagine some aspects of adaptation that are subject to scientific analysis, but it’s not even the science NSF is pushing—it’s people who can do the science.) Then, on top of that, there was the article shared by a Facebook friend that Trump has argued to reduce the overhead rate federal granting agencies pay to institutions to 15% from the roughly 50-70% they pay now.

Because, like the host of this Substack and many of its readers, I am an academic, I am steeped in the controversies over DEI, government pauses in funding, reductions in overhead, etc. Fortunately, most of it doesn’t affect me because I’m semi-retired—I focus on research and service. I fund my own research and, in the service I perform, I studiously avoid these topics. I’m proud of the fact that part of my service is to run a student research grant program, and I emphasize nothing but merit in awarding those grants. I have a Facebook page, and I follow many of my academic colleagues. Ever since Donald Trump was inaugurated and started doing stuff (imagine, a president who does stuff), they’ve been losing their minds. It’s almost gotten to the point that, between their posts and all the ads, I should abandon Facebook altogether, and I would if it weren’t for the trickle of good stuff from my family and those friends who are not academics, fortunately a fairly large group because I have a life outside of academia (like flying).

One of the biggest sources of panic among my academic friends is DOGE. I’m sure you’re familiar with the rhetoric: DOGE is endangering personal information, why should a bunch of wet-behind-the-ears children have access to this information, etc. etc. The part that has baffled me is that people are making this out to be a personal threat. That and the argument that they are unaccountable and unelected, which overlooks the fact that they’re accountable to the president, who was elected.

Let’s look at some numbers: There are roughly 330 million people, including Donald Trump, Ben Shapiro, JD Vance, a host of blatantly racist people, as well as progressives and Democrats, and many Substack readers, who have Social Security numbers. Given the extremely streamlined staff of DOGE, reported to be as few as five, excluding Elon Musk, the chance that they are spending their time burrowing into the life details of any individual is not significantly different from nil. As has been widely reported, they are using AI to match payments with other information to try to identify payments that are waste, fraud, and abuse (WFA). That is something AI can do far, far more efficiently than any human, let alone a small (or even a large) team of humans, without compromising anyone’s data. I heard one congresscritter (D, somewhere) complain that “Big Tech” (read, Elon Musk) should not have access to our Social Security numbers. Um…I hate to tell you this, but it already does. Big Tech doesn’t make it easy--just for grins and because it would be no harm, no foul, I entered my late husband’s Social Security number into the Google search engine and came up with nothing, not even his Social Security death record—but if you’ve filed taxes online, applied for a replacement birth certificate online, or done anything else online that requires use of your Social Security number, it’s out there. Heck, you now have to log in online just to access your own Social Security information.

The kind of data matching that DOGE is doing is something government should have been doing all along. I can almost tolerate the counterargument that there are simply too many payments to too many people and organizations to track the WFA. The sheer weight of the problem might be part of the reason government has had a hard time doing so. If that’s the case, it’s still not an excuse, and now that AI has actually made it possible to check for WFA on a large scale, everyone should be on board.

There was a time when Democrats would have enthusiastically agreed. After all, with the exception of Bill Clinton (whose actions as president eerily mirrored some of Trump’s but without the technical help), expanding the welfare state has been more a Democrat goal than a Republican one (although both are guilty), and presumably, they don’t countenance WFA; the less WFA, the more money available to help real people. Everyone agrees the government spends too much; that there is WFA (which department is worse depends on your politics); and that this WFA should be stopped. I, for one, welcome the ability of AI to finally tackle this problem.

My friends and family used to tease me because I would often pinch pennies on small purchases. My answer was always, it’s because I do that I can afford the large ones. Almost all instances of WFA of taxpayer money are, individually, microscopic compared to the total federal budget. The outcry against shutting down USAID has included this argument. But put all the specious payments together and the scale is staggering. Putting one’s head in the sand to deny it’s happening on a large scale doesn’t solve the problem, and where such WFA has been exposed it has turned out to be eye-watering. For example, Washington State alone paid out $647M in fraudulent COVID claims. Yes, they recovered about two-thirds of that, but how much more money was spent trying to recover those funds? Better to not have paid the claims in the first place. And that was just one state and one program. Multiply that by the hundreds or thousands of programs to which our taxpayer money goes and, well, you do the math.

Finally, another complaint about DOGE is that it’s not transparent. Seems to me it’s been very transparent. A lot of stuff goes on in the government that we never hear about and, to listen to some, is actively hidden from the taxpayers, revealed only if there are whistleblowers, who know they’re putting their careers on the line, despite the so-called protections. In contrast, we know exactly what DOGE is doing.

Now for the overhead reduction. Some might remember that the transfer of money to universities as part of most federal grants started as a result of Sputnik. Sputnik scared the heck out of the US because it showed that the USSR was far more technologically advanced than we knew. (Because I was only 6 years old, my main memory of this momentous event is seeing Sputnik pass overhead my house one night. My dad brought us out to watch; as an aeronautical engineer, he was probably enthralled and appalled in equal measure). The government wanted the US to catch up, so it started pouring millions into universities and national labs to bring us up to speed, an effort that was spectacularly successful.

Because of this, it’s easy to argue that, given Trump’s (and Musk’s) interest in keeping the US in the forefront of technology and science, we should definitely continue payments of overhead to universities. How overhead is calculated and spent is a very complicated topic that is outside the scope of this column. But a significant portion of overhead is devoted to administrative costs, and those have only increased, not least because of the growing (and ever intrusive¹) compliance costs. The only numbers I could find for government regulations of research in university went just to 2014, but between 1992 and that year, the number of regulations increased from 2 to 92. Compliance, which has become increasingly complicated (as any researcher with a federal grant knows), requires staffing—someone has to oversee all that. If the overhead cuts are accompanied by deregulation, something Trump is known for, maybe 15% would be enough. People, after all, are expensive. The arguments for maintaining the current rates are going to have to be a lot more convincing than “we’ll have to fire people”.

In summary, maybe I’m just naïve or (more likely) old and crotchety, but I can’t seem to wrap my head around any program that doesn’t involve merit, something that is of great interest to readers of this Substack. Merit includes how the government spends our tax dollars. I would love to see more money for meritorious research. Maybe if we got rid of all the WFA and the focus on anything but merit, the scientific and technological power of the US would be truly unleashed.

1. In the course of researching this essay, I ran across a fascinating article that laid out all the problems universities face. This article could have been written last year, but wasn’t: Kaplan, N., 1960, Research overhead and universities: Science, v. 132, p. 400-404.