Fresh Blood

If you're the new General, do you consult with the people who have not advanced or do you rely on your own initiative?

Jennifer Pahlka and Andrew Greenway have written a report on government reform. They have been in the trenches of government reform and digitization for years now. They have written a report for the Niskanen Center on the nature of the problems of government implementation. I hope that DOGE is talking to people like them.

It starts out strong and sympathetic.

“Our administrative state’s structures, processes, and ways of working are simply no longer fit for the jobs we need them to do.” Elon and Vivek, we agree with you. Really, we’re on your side. You should listen to us. We’re smart, too.

“Government can and must be reorganized to meet the current moment. This requires prioritizing the government’s ability to achieve its policy goals – what academics call “state capacity” – in a way not seen since the early 1900s. It’s time.” This jargon indicates that this conversation is a safe space for academics and bureaucrats, too. We understand that the role of government is not to signal virtue but to obtain the noble outcomes the American people demand. The failure to make this happen has caused them to lose trust in government.

“But make no mistake: whether or not we do it [rebuilding state capacity] is not a technocratic choice. It’s a political choice – a leadership choice.” This is going to involve spending significant amounts of political capital; that’s why it hasn’t happened to date. For this to work, DOGE will have to generate its own political capital, bootstrapping its credibility one success at a time. You need quick wins. Start small and pyramid into larger success.

“Many [civil servants] are comfortable with the status quo, for sure, but a growing band is hungry for permission to reinvent the bureaucracy in pursuit of the mission that inspired them to serve in the first place. These public sector patriots want to make the government work better, to have a firm and capable grip on the problems it is charge with solving. They are an enormous national resource; the incoming leadership would do well to recognize them and leverage their capacity to the fullest extent.” You should be careful in wielding the axe. You need these people more than you appear to realize. Take the time to figure out who’s on your side and who isn’t and then make your cuts, otherwise you’ll lose vital institutional knowledge without which your project is imperiled.

Pahlka and Greenway start with the Progressive movement, something they label as “America’s last major step forward in government reform.” We want to anchor the discussion.

“Institutionally, the Progressives logged many successes, but they are also in part responsible for making American government more complicated: They had a penchant for layering new structures on top of the old in order to work around corruption and inefficiency. We suffer mightily today from the accumulation of these layers. And while important pro-integration reformers counted themselves as Progressive, the movement was also deeply intertwined with racism and segregation.” The Progressive movement was a fix for the corruption that preceded the professionalization of the civil service. By referring to its racism (and everything was tinged with racism back then, so this criticism is redundant and implicit), we give ourselves license to criticize it with certain politically correct constituencies.

“The Progressives fought to rebuild government fit for the Industrial Era.” If you pay attention to the subtext of what we’re saying, then we need to rebuild government for a new era of Digitization and Artificial Intelligence. New eras have consequences for culture and industrial organization. Put another way, we can compliment our Progressive forebears (for our politically aligned friends on the left) while advocating change. Instead of learning in a linear way, we need to adopt the non-linear techniques of the entrepreneur: test, learn, adapt, test some more, etc. We are on your side. Seriously. You should believe us.

“American bureaucracy is vast and heterogenous, and solving a performance problem within any agency requires a deep dive into the particulars of the situation. Bureaucratic failure arises from a complex interaction of rules and culture, and anyone hoping to fix it must appreciate both. But certain patterns of dysfunction arise time and again.” You need people like us because the problem is far more complex than your public comments suggest you understand.

“What are the forces making the government so slow? The first of those dysfunctions is what Nicholas Bagley of the University of Michigan calls the “procedure fetish,” and we dub the bureaucratic anxiety cycle. Anxiety about legitimacy and accountability drives critics to demand, and bureaucrats to seek refuge behind, more and more layers of procedure that show things have been done “by the book.” But all that procedure further erodes both legitimacy and accountability by overburdening the bureaucracy, reducing its ability to deliver meaningful outcomes.” There is game theory to consider. Agencies know they need to make things happen, but individual bureaucrats cover themselves individually with an expanding array of protections, ostensibly for the benefit of the agency’s reputation. This produces a doom loop in which poor service begets complaints leading to more rules that in turn bedevil performance. This is part of the culture.

“This downward spiral is exacerbated by a complementary anxiety cycle in the legislative branch. Legislators are often frustrated with the performance of the administrative agencies they oversee. They frequently respond by being ever more detailed and specific in their instructions, reducing the agencies’ sense of responsibility and accountability to deliver outcomes.” Seeing the inability of the agencies to accomplish their objectives, the legislative branch imposes more constraints on the decision-making process. Everyone is worried about looking bad.

“Another dynamic that causes policy accretion is what we call the tyranny of tiny decisions. Whenever a process is created in any large organization, there are always a series of suggestions for additional qualifiers and considerations that are individually proportionate and sensible responses to a possible risk. The problem is that when you look back and add up all the tiny, sensible things, you have created a clunking behemoth.” More is not better. But you knew that.

“The constant accretion of policy and process has another source: adversarial legalism. This, too, becomes a cycle. When government is sued or an agency called before a Congressional committee, it generally defends itself on the basis of having followed the proper procedure. But parties seeking advantage find in those procedures opportunities to attack or at least slow down government action. The attacks spur bureaucracies towards ever-more detailed, voluminous procedures, repeating the cycle.” Activists have discovered how to manipulate the process. Don’t think you’re invulnerable to this. They’re going to hit you hard.

“One reason policymakers often misjudge the burden their guardrails will impose is a further dynamic we call the cascade of rigidity. The cascade begins with high-level principles outlined in legislation and executive orders. As these principles descend through layers of bureaucracy, they are translated into more specific and prescriptive guidance. Each level of government, from agencies and sub agencies to individual bureaus and divisions, interprets and operationalizes the guidance in its own way.” The road to hell is paved with good intentions. Everyone gets a say at every level. Legislative and executive intent dissolves in this stew of administrative reagents.

“We also find ourselves somewhere other than where policy intended because we are trying to solve complex problems with tools suited for merely complicated ones.” Complicated problems are ones that you can solve with rules and processes. Complex problems have too many interdependencies and unknown factors for such an approach to succeed. When you have a hammer, everything looks like a nail. Entrepreneurial experimentation is the only way to address complexity because it surfaces these additional dimensions. We’re just like you, really.

“In the technology world, there is a role designated to make those tradeoffs, and it’s telling that it has not until recently existed in government. That job is called a product manager. If you say “product manager” in policy circles, people hear “project manager,” a prolific title across government implementation teams. But the two are different. Project management is the art of getting things done. Product management is the art of deciding what to do. Project management is critically important, especially in government where there are so many things to do. But that list of things to get done is often the superset of wishlists (at the project level, usually expressed as requirements) from a wide variety of stakeholders. If there’s no product manager doing the work of deciding what’s in or out, focused on the desired outcomes instead of checking hundreds of boxes, all the project management in the world will not result in a usable product. On the contrary, more project management means more bandwidth to add toppings to the bagel, until it becomes inedible.” If it wasn’t clear that this letter is addressed to Musk and Ramaswamy, we want to make it explicit that we are technology people first. You need to understand the mindset. We are presenting it in terms with which you are familiar. The bigger question is, why isn’t there a product manager role?

“Over-reliance on contracting exacerbates all the other dysfunctions we discussed above. In an environment of procedural bloat, contracting adds even more procedures. It adds even more layers in the hierarchy, meaning the cascade of rigidity has even more steps through which the intent of policy can be perverted. It adds even more temporal, organizational, cultural, and structural distance between policy and implementation teams, making it far harder to engage in test-and-learn cycles. Software that’s contracted out often can’t be modified by government staff, making it hard to react when needs change.” Outsourcing to contractors often exacerbates the bureaucracy, particularly when government agencies lack internal teams with the skill sets to outsource well.

The article then lays out its solutions in four categories.

“1. Hire the right people and fire the wrong ones 2. Reduce procedural bloat 3. Invest in digital infrastructure, and 4. Close the feedback loop between policy and implementation.”

You can imagine what they say. It’s a good read.

Is anyone in DOGE paying attention? Or will they wade into the problem ignoring the advice of those who have been in the trenches precisely because the trench warfare hasn’t made any historical progress?