DOGE, Software, and Government
a truly impossible mission
The core work of DOGE will be less about making minor improvements to existing IT systems than in overhauling the federal government’s tech-stack from top to bottom. If successful, DOGE will bring unprecedented transparency to federal spending while laying the infrastructure needed for a significant downsizing in the federal workforce through automations.
For example, the DOGE is reportedly planning to move the systems used to track federal spending onto a permissionless blockchain. Curious where your infrastructure dollars are being spent? Today, accessing that information may require combing through obscure PDFs and press releases or submitting a FOIA request. In the near future, it could be as simple as making an API call to fetch the Department of Transportation’s on-chain transactions.
DOGE has become the great Wishing Well of political punditry.
It was supposed to find enough savings to dramatically reduce the Budget deficit.
Another fantasy is that it will cut through the thicket of regulations.
And now, it is going to modernize the software infrastructure of the government.
The first mission is made impossible by entitlements and Congress. The second mission would get stuck in legal quicksand. But the third mission is the most impossible of all.
The software mission runs counter to one of my aphorisms (which I forgot to include in Klassic Klingisms):
Every Organization Gets the Information Systems it Deserves
I noticed this when I was at Freddie Mac in the early 1990s. Freddie Mac buys mortgages and sells securities backed by those mortgages. The securities were like bonds, with Freddie Mac paying the security holders every month, based on principal and interest received from the mortgages. If a mortgage defaulted, Freddie Mac paid the security holders the full value of the principal.
There were some departments that were highly disciplined. They operated according to well-articulated procedures. They produced reliable results. The department that managed the payments to security holders was like this. Naturally, the process was computerized.
Other departments were much looser. The marketing department, which negotiated contracts to buy mortgages from lenders, did not much care to have any limits placed on its creativity. Salesmen would put together agreements that stipulated how lenders would have to share in the risk of subsequent mortgage defaults, but the agreements were stuck in a drawer. By the time defaults occurred, the cost-sharing agreements might be forgotten. For this undisciplined department, there was little or no computer support.
Freddie Mac’s Information Technology division would periodically launch an initiative to rationalize all of the company’s computer systems. But the business departments resisted, and the loosest departments—the ones that needed information engineering the most—provided the least cooperation. Colleagues pointed out that having the systems initiative come from the IT department was like trying to drive a car from the back seat.
Electronic Medical Records
In my 2006 book on health care, Crisis of Abundance, I pointed out that our health care system is very fragmented. At the time, some technocrats had high hopes that electronic medical records and other information systems would lower costs in the health care system. But the process of delivering health care is very unsystematic, and you cannot fix an unsystematic process by throwing computer resources at it.
And lo and behold, the hope of using electronic medical records to reform health care still lives.
Liberate healthcare data - allow all patients and healthcare providers to benefit from seamless sharing of health records;
My point is that you cannot do this until you tighten up the health care delivery process, making it more rigid and uniform. And I would not try to do that. Health care does not necessarily lend itself to being commoditized. You risk making health care in America less open to innovation and less responsive to the needs of people.
So far, all that has been accomplished by the electronic medical records drive has been to put small physician practices out of business. They have not been able to absorb the overhead involved in implementing these systems, so that they have been forced to lose their independence, primarily to hospital-owned conglomerates.
As an aside, I think that the other ideas that Joe Lonsdale’s partner has for trying to improve the cost-effectiveness of health care in America also miss the mark. If you change the metrics for compensating health care providers, in the end all you do is give them a different system to be gamed. What you have to do is reduce the incentive for consumers to obtain procedures that have high costs and low benefits.
We are not Estonia
Which brings me to DOGE and government. Hammond points to Estonia as an example.
Take Estonia, which has the most most sophisticated e-government in the world. Given the benefit of a blank slate following the collapse of the USSR, Estonia’s e-government was the product of young policy hackers who entered the civil service when the internet was just starting to take off. With the foresight to see where technology was heading, these reformers pushed Estonia to become digitally-native through a series of sweeping modernization initiatives.
The key phrase here is “blank slate.” We don’t have that here. What we have are hundreds of agencies with long histories, overlapping missions, very poor cross-communication, and no motivation to cooperate with one another or with the President. There is zero chance that any broad information-technology initiative can succeed.
I have suggested that we need a major re-organization of the executive branch. I also recommended a Chief Operating Officer, along with a very powerful Chief Auditor. Something along those lines is a pre-condition for any information systems effort that is supposed to cut across the entire executive branch. If Elon’s mission is to try to improve government by focusing on software, leaving the crazy-quilt structure of the executive branch in place, then I would advise him not to waste another minute on DOGE.
substacks referenced above:
@
When it comes to government and IT, my fear is less about efficiency than it is about how easily it seems to be that our adversaries are able to exploit rotting systems.
I think the comparison to healthcare is on point, as the single payer advocates on the left have no understanding of the technical and process debt that will not allow you to just wave a magic wand to bring about utopia.
I think part of the impossibility of upgrading the IT infrastructure is simply resistance to change. Certain stakeholders certainly like the status quo, for a variety of reasons (from relatively benign to borderline malicious). But the “this is The Way we’ve done it” mentality is strong.
I hired on with a DOE prime contractor a little over two years ago. When first shown some of the systems we have to use, my thought was “you’ve got to be kidding me.” The systems were obsolete well over a decade prior.