Gregg Tahmisian
March 17, 2025
Change is happening in 2025, whether we like it or not. In its effort to cut costs within the Federal government President Trump’s Department of Government Efficiency (DOGE) is reducing the size of the government through layoffs and, in places, ripping out entire departments. DOGE took over the United States Digital Service (USDS), renaming it to the United States DOGE Service. The new USDS began establishing DOGE teams within the federal agencies to carry out spending cuts and “modernize federal technology and software to maximize governmental efficiency and productivity.” In the process, other federal groups with similar missions became redundant.
One such group was 18F. Operating within the General Services Administration (GSA), 18F aimed to help federal agencies build and acquire better software. DOGE “deleted” 18F entirely. The Technology Modernization Fund (TMF), a unit also within GSA that supports federal modernization efforts through loans and project management, may face a similar fate, though perhaps at the hands of lawmakers rather than DOGE.
These and similar cuts coincide with increased reporting on fraud, waste, and abuse caused by – or at least abetted by – the antiquated computer systems upon which federal agencies often rely. This combination leads many to worry that the President is shooting himself in the foot, so to speak, by eliminating the very people that can best modernize government technology to reduce the fraud, waste, and abuse that he wants to see eliminated.
Here’s the problem with that thinking: The previous solutions haven’t worked. The ROI across the board has been unacceptably low. The well-known definition of insanity is doing the same thing over and over and expecting a different result. Well, DOGE is not going to allow “the same thing” to continue. The glaring truth is that USDS, 18F, TMF, and other such initiatives, have consumed billions of dollars and have failed to reliver an acceptable return on that investment.
Which begs the question, “Why not?” From my perspective, it boils down to two key areas: the first is what I call the government-helping-government (GHG) effect; the second is prioritization of monolithic, or “blob” software over marketplace platform solutions. The common feature of both is that they limit private sector competition, which leads to higher costs and worse service across the board.
First, the GHG phenomenon is basically when government agencies morph into a version of a contractor, consultant, or provider to other areas of government. The problem is that, for several reasons, government is inherently not good at providing these types of services, and they inevitably fail to deliver.
Initially, the GHG proposition seems appealing, and it can start out with good intentions. For example, when an agency achieves a technology success, they want to share it with other agencies in order to help them be successful as well. Who wouldn’t want that, right?
So, the agency begins to build out the infrastructure to provide the technology as a service to other agencies. They need to promote the service so that other agencies know about it. They need a Project Management Office to “onboard” these other agencies. They need technologists to support the other agencies. Of course, all this needs to be staffed and paid for. And before you know it, you have a full-fledged GHG “vendor” embedded in an agency – a vendor that is financed by public dollars. Additionally, because of government procurement rules, it is almost always easier to “buy” a service from another government agency than it is from the private sector. Thus, the GHG crowding-out effect ends up significantly limiting competition in the space. Predictably, less competition means higher costs and worse service.
Second, prioritizing what I’m calling “blob” software happens when agencies look for a software solution that solves a single giant problem. One extreme example might be, “We need a software solution to run the nation’s air traffic control.” (Note: I don’t know whether anyone actually thinks this way about air traffic control software; it’s just an example.) When you put it this way, there are maybe only a couple companies that could conceivably deliver such a complex piece of software.
However, air traffic control is a “blob” of a problem that can be broken down into smaller, well-defined pieces that can communicate with each other to form the whole system. When those pieces are decomposed and sufficiently defined, it opens the way for many smaller, perhaps more specialized, vendors to compete to produce these smaller modules. And if the system is well defined such that different versions of the same module – even when produced by different vendors – are interchangeable, it opens the possibility to have a “store” where the agency can choose from different modules supplied by potentially many different vendors. This is a marketplace platform.
Such a platform allows the agency to choose which vendor supplies each module based on whatever combination of features, service, and price is important to the agency. Additionally, the agency can upgrade modules one-at-a-time rather than having to scrap the entire system and rebuild from the ground up.
This is just applying the principles of the industrial revolution and interchangeable parts. For example, it’s how cars are built: Everything from spark plugs to tires to radios come in well-defined modules that can be independently replaced without having to scrap the car and buy another one. If for example, you wanted dedicated snow tires rather than the all-season tires that came with your car, you can buy them individually from a number of vendors who will compete for your business on price, features, and service. We should expect no less from our software.
It is a tall order, but I am confident that if DOGE can eliminate the GHG effect and bring marketplace platforms to government software, it can dramatically improve governmental efficiency and productivity.
