A Memo to Messieurs Musk and Ramaswamy

There is no doubt that the government has grown bloated and inefficient. Examples have been visible, and frankly risible. As only one example, the Space Launch System is massively behind schedule, massively over budget, never likely to be economical, and doesn't actually work. Still, it has the advantage that it has subcontracts in 45 states. In other words, it has approached budgetary nirvana — subcontracts in all 535 congressional districts. 

If DOGE does nothing else but reduce redundancy, eliminate wasteful programs, and trim unnecessary spending, it will have done noble work and made a significant difference.

I believe DOGE should not limit itself to just that. A second — and frankly, harder — task is to ask the question: Does the government know what it's doing?

I don't mean this in the sense of asking whether the government is incompetent, although it wouldn't be hard to argue. But I think that's an accident. I think the underlying problem is that the government bureaucracies have no effective direction toward a desired result. 

Consider these examples:

• Hurricane Helena Relief where families are still living in tents as snow is falling, while FEMA-provided temporary shelter goes unused and people are being evicted from tiny homes built by volunteers because someone didn't check the right boxes on a building permit.
• Electric Vehicle Charging Stations for which billions of dollars were allocated, but only a handful — variously reported as 8 and 17 — have actually been built.
• Rural and Underserved Internet Access for which $42 billion was appropriated years ago, and through which not one person has received access. (I suspect this is a particular sore point for Mr. Musk.)

The reasons for this are myriad. The operation of bureaucracy introduces expensive inefficiencies in a thousand ways, but I propose the major one is rarely, if ever, mentioned. These bureaucracies don't know what they are doing in a very specific sense: they have no clear goals; there are no effective measures of progress toward those goals; and there are no rewards for achieving those unstated and ill-defined goals.

There have been times in history in which this has not been true. In wartime, the clear goal is to defeat the enemy. (There are also too many examples in which "defeat the enemy" has been subordinated to other goals, but that's a topic for another memo.)

One stellar example of this was the Apollo program. Originally conceived of in the Eisenhower Administration, it was focused by President John F Kennedy in 1961 as a program to put man on the moon and return safely by the end of the decade. Sure enough, even though the original sponsor was murdered in 1963, Neil Armstrong and Buzz Aldrin landed on the moon in July of 1969.

It was a clear goal. Many felt it was an impossible goal, but it was in fact an achievable goal, and there is the famous story of the janitor who was asked, "What do you do here?" He answered, "I'm helping send man to the moon!"

There is a recent management fad called "Objectives and Key Results" (OKR). Like many management fads, it works well if it's applied well, but too often it is just another fad, applied without thought and eventually discarded. It has been very effective when used well at places like Google.

The Apollo program, while it was long before OKR was codified, was an example of an effective project managed by an overall audacious goal, with a well-defined key result. It was decomposed into nearly as audacious intermediate goals: one man in orbit; multiple men in orbit for longer missions; in-orbit rendezvous; successfully sending men to orbit the moon; and finally the actual landing. There were failures, there were accidents, men were lost and mourned, but through it all, "I'm helping send man to the moon!" focused the effort.

And, of course, it was followed by the Shuttle program, which ended up a hodgepodge of goals from many customers — a heavy-lift vehicle for Keyhole reconnaissance satellites, a crew vehicle, a research platform. It became a standard government program.

Related: Elon Musk Wants to Save the World — and More

My proposal is this: While eliminating wasteful programs, adopt an OKR-like approach to everything. If a program is to be retained, it should have a clearly stated goal and a single person on whom responsibility for achieving that goal rests. Have clearly stated measures of success and transparently reveal those measures and the results. Reward the responsible person for achieving the desired results.

This is, I realize, a radical rethinking of current government operations, where we can rarely find a person responsible for achieving goals, and for which goals are too often vague or even completely unstated. (Can you tell me what the purpose — the goal — of the Department of Transportation really is? Can you tell me how we would know if it was achieved?)

But consider a relatively small example: providing internet connectivity to rural and underserved Americans. And yes, in terms of government operations, a mere $42 billion is a relatively small example.

The overall objective is actually pretty simple: provide internet to people who want it. The key result to be measured is how many people are being connected to the internet. And the "Internet Czar", whoever it is, gets rewarded for the number of people who are actually connected to the internet who didn't have an internet connection before. 

As I said, I realize this is a really radical proposal. It might even be a modest proposal. And it would be an ongoing struggle because people in general don't have the mindset to achieve well-stated goals and accept responsibility for achieving them. 

But imagine a government where the American public knows what they're paying for in government and can see if they are actually getting what they pay for.