Same Name

In March 2026, the Office of Personnel Management launched a campaign to recruit Gen Z workers into the federal government. The pitch: “make government cool again.” The context: over the previous year, 386,826 federal employees had departed — a 9% net reduction in the workforce. Among those who left, a third of employees with 30 or more years of experience. Four million years of accumulated service, gone.

The replacements are being hired. The question is what, exactly, they’re replacing.

How Fast Does It Break?

On May 16, 2003, Coalition Provisional Authority Order No. 1 removed all senior Ba’ath Party members from their positions in Iraq’s government. One week later, Order No. 2 dissolved the Iraqi military. In two weeks, two documents eliminated approximately 485,000 people — the entirety of Iraq’s experienced administrative and security apparatus. The party membership that qualified them for those roles was the same membership that now disqualified them.

Twenty-two years later, Iraq’s institutional capacity has not recovered.

Boeing took a different path to the same destination. In 1997, McDonnell Douglas merged with Boeing and installed a financial-metrics culture over what had been, in the incoming president’s words:

“When people say I changed the culture of Boeing, that was the intent, so that it’s run like a business rather than a great engineering firm.”
— Harry Stonecipher, Boeing President, 2004

In 2001, Boeing moved its headquarters from Seattle to Chicago — 500 executives, 2,000 miles from 30,000 engineers. Through the 2000s, the 787 program outsourced 70% of development. Through the 2010s, waves of early retirements drained the experienced engineering workforce. In 2018 and 2019, two 737 MAX crashes killed 346 people. In 2024, a door plug blew out of an Alaska Airlines jet because four bolts were never reinstalled, no one documented the removal, and only one of twenty-four door team members had relevant training.

Two weeks in Iraq. Twenty-five years at Boeing. The timescale differs by three orders of magnitude. The outcome is structurally the same.

Three Layers, One Collapse

What these cases share is a layered structure of institutional knowledge and the asymmetry between how fast it breaks and how slowly it rebuilds.

The surface layer is codified — procedures, manuals, regulations, documented processes. This is what can be written down, and what reconstruction efforts target first. It rebuilds in months.

Below that sits relational knowledge — the informal networks, the inter-agency coordination, the understanding of who knows what and whom to call when the manual is silent. This requires people to have worked together long enough to develop trust. It rebuilds in years.

At the bottom: experiential knowledge — the tacit judgment accumulated over decades, pattern recognition from context, the institutional memory of what went wrong last time. This requires relational networks to have weathered enough crises that the institution develops collective memory. It rebuilds in decades, if it rebuilds at all.

Destruction collapses all three simultaneously. A two-week purge or a twelve-month workforce reduction eliminates the documented procedures, the networks those procedures relied on, and the judgment of the people who wrote them — all at once. Reconstruction can only proceed from the bottom up. You cannot rebuild experiential judgment without first having relational networks, and you cannot rebuild relational networks without first having people in roles long enough to develop them.

Three Ways to Lose It

The cases separate into three erosion mechanisms. De-Ba’athification, the Cultural Revolution, and DOGE represent sudden purges — all layers collapse simultaneously because the people who carried them are removed at once. The post-Soviet brain drain is an emigration drain — slow-drip loss where the relational layer fragments first as key nodes leave, experiential knowledge follows because judgment degrades in isolation, and codified knowledge persists longest because textbooks don’t emigrate. Boeing is deliberate cultural replacement — what the institution rewards is changed, the knowledge-holders leave on their own because their expertise is no longer valued, and the institution is gradually rebuilt around different priorities.

Three mechanisms, same endpoint. The institution that emerges carries the original name.

What Comes Back

China’s Cultural Revolution closed universities for a decade, sent professors to farms, and reduced the country’s scientific output to a single paper in the Science Citation Index in 1973. When Deng Xiaoping restored university entrance exams in 1977, reconstruction began — but what emerged over the following thirty years was a different scientific establishment. More applied, more globally connected, more market-oriented. The Chinese Academy of Sciences persisted in name. The institution was new.

Boeing still builds planes. Iraq still has ministries. Russia’s research institutes still exist. In every case, the institution that inherits the name after a knowledge collapse is functionally different from what preceded it — competent in different dimensions, blind to different contexts, building judgment from a different starting point.

Germany after 1945 is the exception that clarifies the rule. Denazification preserved institutional capacity because it separated the political question from the knowledge question. Top leaders were removed. Everyone else was vetted and reintegrated. No permanent exclusion mechanism was created. The variable that determined the outcome: whether political affiliation and institutional knowledge were separable. In Germany, they were — most party members were also competent administrators. In Iraq, they weren’t — party membership had been the prerequisite for institutional roles.

DOGE sits between these poles. Federal workers are not party members. Their knowledge and their employment are, in principle, separable. The 25,000 rehired as “essential” — and the court-ordered reinstatements at 18 agencies — suggest the system recognized this, belatedly. But 264,000 net positions remain gone. And the Gen Z hiring campaign, whatever its merits, is recruiting for roles where the codified layer can be taught but the experiential layer can only be grown.

Same procedures, new people. Same name, different institution. The new hires will learn the manuals in months. The networks will take years. The judgment — the instinct for which exceptions matter, which precedents hold, which shortcuts kill — will take decades they may or may not get.

Meanwhile, the institution carries on under its original name, performing functions that look similar from the outside. The gap between “someone in the role” and “someone who knows the role” is invisible until the moment it isn’t. Boeing built planes for twenty-five years in that gap. Four bolts were the difference.