PDF Summary:Who is Government?, by Michael Lewis

Most Americans can’t name many federal employees or describe what government workers actually do all day. Michael Lewis—bestselling author of The Big Short and Moneyball—set out to fix this. In Who Is Government?, Lewis and six other writers profile a wide range of federal workers who save lives, outperform private companies, and protect democracy’s foundations by solving problems markets don’t address.

These workers aren’t political appointees but career civil servants, including a mine safety engineer who spent 30 years preventing roof collapses, an FDA specialist who connects doctors with treatments for rare diseases, and a cemetery director who achieves customer satisfaction scores higher than any private business. Along with these profiles, we’ll explore why the private market fails at these tasks, the legal constraints that prevent agencies from telling their own stories, and the cognitive biases that make government work invisible when it does its job well.

(continued)...

Why Participation Makes for Great Publicity

The federal government runs roughly 350 citizen science and crowdsourcing projects across 26 agencies, asking citizens for help with everything from transcribing records to tracking whale migrations. Just as the National Archives can’t transcribe 13 billion records, NOAA-affiliated researchers can’t monitor every whale across thousands of miles of ocean. The solution is the same one Wright hit upon: Turn the public into collaborators. In the North Pacific humpback whale photo-ID dataset, which scientists use to study feeding, breeding, and migration, 34% of all encounters were recorded by citizen scientists. They submit photos via platforms like Happywhale, which uses AI to identify whales by their tail-fluke patterns.

These programs have a side effect that connects to Lewis’s larger argument. When ordinary people volunteer for a federal science project, they don’t just contribute data—they come away with an understanding of what that agency actually does. That personal experience is far harder to dismiss than any statistic about agency budgets. These projects promote a spirit of open government by giving volunteers a direct line of sight into agency missions. When people transcribe census records or photograph whales, they’re not just helping solve a resource problem. They’re becoming the kind of informed constituency that Lewis argues is missing: citizens who understand firsthand how crucial government work really is.

Why Does Government Work Matter?

Lewis and his contributors argue that the work federal employees do is profoundly consequential—that without the staff of many government agencies and departments, essential work would go undone. In this section, we’ll explore the specific ways this work matters: It saves lives that would otherwise be lost, it achieves excellence the private sector wouldn’t sustain, and it protects the factual foundation that democratic governance requires. We’ll also examine why the authors believe it’s particularly urgent to recognize the value of federal service in the mid-2020s’ political climate.

Government Work Saves Lives

First, federal employees do work that saves people’s lives. Heather Stone, whose FDA work on rare diseases we discussed earlier, exemplifies this life-saving role. Stone created CURE ID to address a gap: Peer-reviewed journals won’t publish case reports about individual patients, yet for very rare diseases, individual cases are the only evidence available. No private entity will build infrastructure for sharing this information because doing so generates no revenue stream. Lewis argues that patients with conditions too rare to generate profit would die without government workers facilitating access to treatments that markets won’t develop.

Stone’s work saved Alaina Smith, a five-year-old dying from balamuthia, a brain-eating amoeba that kills ninety-five percent of those it infects. After four months of conventional treatment that appeared to be killing her faster than the amoeba, Alaina’s grandmother found a research paper mentioning Stone—she’d helped another doctor obtain emergency FDA authorization for nitroxoline, an antibiotic not approved in the US but effective against balamuthia. The family tracked Stone down, and within 24 hours, she’d secured emergency authorization and located unused pills from a California patient. Just days after starting treatment with nitroxoline, Alaina’s symptoms vanished. Two years later, she appeared completely cured.

Shortly thereafter, a four-year-old in Northern California died from balamuthia because her doctors learned about nitroxoline too late. The difference between life and death came down to whether a worker like Stone could connect doctors with information that has no commercial home.

Why No Single Fix Can Solve the Rare Disease Problem

Rare diseases are individually uncommon but collectively widespread: 30 million Americans live with one of more than 7,000 identified rare conditions. As few as 5% of these diseases have an FDA-approved therapy. The problem is so entrenched—spanning how drugs are developed, how doctors are trained, and how information is published—that the government has built a patchwork of solutions, each addressing one dimension of a failure the market can’t repair.

The Orphan Drug Act of 1983 was designed to solve part of this market failure: Diseases affecting fewer than 200,000 Americans didn’t promise enough revenue to justify the cost of drug development, so the law created financial incentives—extended market exclusivity, tax breaks on clinical trial costs, and waived application fees—to change that calculation. The decade before the Act produced just 34 treatments for rare conditions, and the tally now stands near 800 approved orphan drugs. But the financial incentives work best for conditions with patient populations large enough to generate substantial returns. The rarest of the rare diseases, in contrast, remain largely untouched.

Even when a treatment does exist for a condition common enough to be studied, the path from published evidence to a patient’s bedside is agonizingly slow. Clinical research takes roughly 17 years to become standard practice. For ultra-rare diseases like balamuthia, the situation is worse: There are no large clinical trials to draw from, and the only knowledge that exists is scattered across case reports and researchers’ personal experience. Medical training compounds the problem. The principle at the heart of diagnostic reasoning(“start with what’s most likely”) means that rare conditions tend to fall outside physicians’ instincts. Yet the math suggests that most doctors will encounter at least one rare-disease patient in their careers.

Stone chose to address the dissemination side of the problem. When no approved treatment exists, physicians improvise—repurposing existing drugs, adjusting dosages, and trying combinations that no trial has tested. But these experiments happen in isolation unless someone builds a system to share them, which is what CURE ID was designed to do.

Government Work Can Outperform the Private Sector

Lewis and his coauthors also argue that government workers can commit to missions and achieve excellence the private sector can’t sustain because profit requirements make it impossible. To illustrate this, Casey Cep profiles Ronald Walters, who leads the National Cemetery Administration (NCA). Any military veteran or their spouse can be buried in a national cemetery free of charge—the NCA operates 155 cemeteries, conducts more than 140,000 burials annually, and maintains four million graves. Under Walters, the NCA scores 97 on the American Customer Satisfaction Index, higher than any private company. This excellence emerges from Walters’s belief that those who served in the armed forces deserve perfection.

Walters spent two decades developing a 40-page manual specifying everything from grass height around headstones to how quickly graves should settle. He created a national call center so families can make burial arrangements immediately rather than waiting for business hours. He launched an apprenticeship program giving homeless veterans jobs in cemetery maintenance. Grieving families notice the immaculate grounds, the speed and ease of making arrangements, and the sense that their loved one is being honored.

When Work Feels Like a Moral Duty

Walters’s insistence on perfection—and the extraordinary results it produces—reflects what happens when a worker experiences their job not as employment but as a moral obligation. People generally relate to their work in one of three ways: as a means to a paycheck, as a vehicle for advancement and status, or as something closer to a personal mission. People who experience that third relationship with their work tend to reshape their jobs, taking on tasks that weren’t assigned, building connections with the people their work affects, and expanding their role until it matches the significance they feel it should have. How they relate to their work has little to do with what’s on their business card or in their bank account.

Walters exemplifies the “personal mission” category, and his emphasis on perfection as something that veterans deserve recalls a finding from a study of hospital cleaners. Those who saw their work as a personal mission didn’t describe themselves as custodial staff at all. They learned which cleaning chemicals might irritate specific patients, arranged their schedules around patient needs, and formed relationships with families—reinventing a job description that said nothing about any of this. But that devotion comes at a cost. Workers with the strongest sense of calling are also the most willing to accept lower wages, give up personal time, and endure difficult working conditions.

Cep notes that the level of care Walters deems necessary at each of the national cemeteries would make no sense for a private cemetery: Cutting corners would increase margins. The reason the NCA outperforms private companies is optimization for different goals. Private cemeteries must satisfy shareholders by maximizing profit. The NCA optimizes purely for service quality. Walters can maintain standards that would bankrupt a private operator because his mission matters more than extracting a profit.

(Shortform note: Research suggests that the NCA does excellent work not simply because it’s free from profit pressure, but because the absence of profit pressure helps attract and retain people like Walters. When federal employees were asked about what keeps them in their jobs, money was far from the top of the list: Factors like the meaning of their work and the chance to contribute to something larger consistently outranked cash bonuses. Plus, when public-sector pay rises to compete with private-sector salaries, the people most drawn to government work for its own sake become less likely to choose it. Balancing government compensation with the private sector’s could filter out the people whose motivation makes them most effective.)

Government Work Protects Democracy’s Foundations

Finally, the authors contend that government work protects democracy by creating the shared factual foundation that collective decision-making requires. John Lanchester profiles the Consumer Price Index (CPI)—which tracks changes in the price of consumer goods—as an example of the government creating this shared foundation. The CPI measures inflation, which is difficult to do since inflation affects everyone but no individual can perceive it accurately. We have to understand and track inflation for several reasons: Social Security recipients depend on it for cost-of-living adjustments, the government adjusts tax brackets based on it, unions negotiate labor contracts using it, and countless private agreements reference it as a benchmark.

Without a neutral measure of inflation, any group could claim an inflation rate that matched whatever benefited them, making economic coordination impossible. Lanchester explains that only the government can create this measure because it requires three things private markets can’t provide. First is neutrality: Any private company has incentive to report numbers serving its interests. Second is infrastructure: Bureau of Labor Statistics workers visit stores nationwide tracking prices for thousands of products continuously. Third is universal access: The CPI serves everyone equally with no customer to sell to and no revenue stream.

Why Statistics Like the CPI Are Essential—and Counterintuitive

The CPI rests on the mathematical idea that that combining many imperfect observations into a single number—an average—can yield more reliable information than any individual observation. Stephen Stigler explains in The Seven Pillars of Statistical Wisdom that this idea, called aggregation, seems to throw away what matters most (the detail of each individual case) to reveal a more accurate picture of the whole. Before averaging took hold, the scientific instinct ran in the opposite direction: If you had five measurements and they disagreed, the goal was to figure out which one was right, not to mash them all together. Combining your best observation with inferior attempts would have seemed perverse.

Modern statistics is built on the idea that the group reveals what no individual can, an idea that’s never been comfortable for some. It took astronomer-turned-social-scientist Adolphe Quetelet to make that leap. Astronomers had long averaged repeat measurements of the same star. In the 1830s, Quetelet decided to average across different people, treating the variation between soldiers’ chest sizes the same way astronomers treated the wobble in their telescopes. His concept of “l’homme moyen,” the “average man,” established the intellectual tradition that would eventually produce tools like the CPI.

This history also illuminates why Lanchester contends that only the government can do the work of creating tools like the CPI. Official statistics are an example of a public good: Everyone benefits from having a neutral measure of inflation, but because no one can be charged for using it, no private company has an incentive to produce it. This is the economic logic behind the three requirements of neutrality, infrastructure, and universal access that Lanchester cites as essential for an official measure of inflation—he’s describing the characteristics of a public good that the market systematically underprovides.

Many Americans are suspicious of statistics like the CPI. When officials announce falling inflation rates, people think that means that prices are decreasing when it’s only the rate of increase that’s slowing: Prices continue to rise, just less rapidly. This disconnect makes people think the statistics are fabricated. But Lanchester argues this suspicion represents a dangerous rejection of Enlightenment ideals about discoverable truth. We use mathematical tools like statistics to uncover patterns that exist independently of our individual perceptions. When people dismiss numerical measures like the CPI, they’re rejecting the idea that we can observe truths larger than our personal experience.

Lanchester warns that this threatens democracy itself: If citizens can’t agree on basic facts, collective decision-making becomes impossible, and society fragments into competing realities where rational deliberation can’t occur.

Why Inflation Feels Different Than What the Data Says

Lanchester argues that nobody can accurately perceive inflation, and research helps explain why. When people estimate how quickly prices are rising, they lean on those they see most often, which tend to be groceries and fuel: You see the number at the gas station every time you drive past it, and you register your total at the grocery store each week. But you probably don’t think as often about how the cost of your streaming subscriptions or health insurance has shifted. This unevenness matters, since everyday expenses make up a much smaller share of consumer spending than we assume. So when grocery prices spike, our mental picture of inflation runs ahead of what the index reports.

On top of that, our memories of prices are lopsided. Rising prices register more strongly than falling ones when people form their sense of how much things cost. The price of a carton of eggs going up by a dollar makes a bigger impression than the price of a bag of apples going down by the same amount.

These aren’t personal failings—they’re predictable features of how attention works, and becoming aware of them can help correct them. This also means you’re not wrong if you feel that the CPI doesn’t match your experience; it only becomes a problem when people use their observations to dismiss the statistics, instead of asking what the numbers mean and why they don’t feel right. Jonathan Rauch argues in The Constitution of Knowledge that societies need more than a political constitution to function. They also need norms and institutions that convert our perceptions and disagreements into facts we can collectively act on. The CPI is one piece of that infrastructure: Without it, inflation is whatever anyone says it is.

Why Government Work Warrants Recognition and Protection

Lewis and his contributors argue that it’s urgent to recognize the value of government work because political attacks on government agencies destroy knowledge, tools, and relationships that took decades to build. Lewis argues that the invisibility of their work enables these attacks, and the stereotype of the government as wasteful fuels the destruction. Because Americans experience the government’s benefits without connecting them to anyone’s effort, there’s no constituency to defend federal workers when they’re attacked. Politicians can eliminate them while claiming to cut waste, and the consequences—destroyed services, lost expertise, preventable deaths—often don’t become apparent until it’s too late to reverse course.

Lewis argues the fiscal rationale for mass workforce reductions (that cutting employees will meaningfully reduce spending) doesn’t align with budgetary reality. The size of the federal civilian workforce has remained stable at about 2.3 million employees for over 50 years, even as the country’s population has grown significantly. These workers represent under two percent of the American workforce, down from 5.2% in 1952. Moreover, federal employee salaries constitute only 5% of total federal spending, with the vast majority going to defense, interest on debt, and programs like Social Security and Medicare.

Nevertheless, as the book went to press in early 2025, mass federal workforce reductions were underway. Stone’s CURE ID website displayed a message that appeared across dozens of government sites: “This repository is under review for potential modification in compliance with Administration directives.”

The Gap Between Appreciation and Protection

Even government work that most Americans appreciate can come under attack. For example, 17,000 public libraries across the US receive more than 1.3 billion visits annually, and most voters say public libraries play an important role in their communities. But appreciation doesn’t always translate into action. A 2016 Pew Research Center survey found that 66% of Americans believed closing their local public library would harm their community, while just 33% said it would make a big difference to them or their family. More recently, a 2024 YouGov survey found that 85% of Americans hold a favorable opinion of libraries, and 47% think public libraries need more funding.

However, the gap between recognizing that something matters in general and feeling it matters to you has consequences: Nearly 90% of library funding comes from local government through property taxes, municipal budgets, or county levies, meaning libraries survive or fail based on whether voters and their elected officials prioritize them at budget time. When a library’s future depends on a special election, turnout can be vanishingly low—as when an Arkansas library funding dispute drew a very small turnout to the polls.

Libraries have also been caught in the crossfire of federal efforts to reduce spending. In March 2025, the Department of Government Efficiency (DOGE) removed Institute of Museum and Library Services (IMLS) staff from their offices and placed almost the entire staff on leave. Those 75 employees had overseen the only federal agency dedicated to library funding, one that provided $266.7 million in 2024, just 0.003% of the federal budget. The cuts proved especially disruptive to services in rural communities, and the elimination of expert positions further eroded effective support for public libraries. Courts eventually intervened, and by December 2025, IMLS reinstated all grants.

But the disruption proved Lewis’s point about institutional fragility. In Massachusetts, for instance, the funding freeze forced the Board of Library Commissioners to slash statewide research databases for students, pull back accessibility services for blind and disabled residents, and cancel improvement grants to 18 local library systems. Even after funding was restored, the process of rebuilding was complicated, demonstrating that the relationships, expertise, and institutional knowledge that make government work effective are far easier to disrupt than they are to rebuild.

A Small Workforce With an Enormous Reach

The federal civilian workforce—roughly 2 million people, or less than 2% of all US workers—administers programs of staggering scale. In 2019, 10 major safety-net programs reached 99 million Americans, nearly a third of the population and including close to half of all US children. The Supplemental Nutrition Assistance Program (SNAP) alone served an average of 41.7 million people per month in 2024, while Social Security paid benefits to more than 50 million retirees plus millions of disabled workers, dependents, and survivors.

Yet as this workload has grown, the workforce delivering all of this is not only small but shrinking in relative terms. The US population has grown by more than half since the mid-1970s, but the number of federal employees has barely budged. And despite perceptions of a bloated bureaucracy, the government actually spends nearly three times as much on private contractors ($759 billion in 2024) as on its entire civilian workforce ($293 billion). That disproportion, as Lewis notes, turns the fiscal argument for cutting the government workforce on its head: Shrinking the civil service targets the smaller, cheaper part of the labor equation and risks increasing dependence on the larger, more expensive part.

How Should Americans Respond?

Lewis and his contributors believe the solution to protecting government work lies in changing public perception through storytelling—a project that requires journalists, writers, and media organizations to amplify these workers’ stories. They argue that telling individual stories can challenge stereotypes in ways policy arguments can’t. When readers learn about specific people doing specific work, they’re less likely to believe the caricature of lazy federal workers wasting taxpayers’ money. The Washington Post series “Who is Government?” attracted four times the expected readership. Lewis interprets this response as a genuine hunger for information that helps people understand what government workers actually do.

How Stories Change Minds

The idea that stories change minds more effectively than policy arguments is supported by cognitive science. Psychologists have a name for what happens when a story grips you: narrative transportation. The term captures the feeling of being pulled out of your own life and into someone else’s. This makes a story persuasive through a mechanism called “reduced counterarguing.”

When you read an opinion article arguing that government workers are valuable, you evaluate its claims critically, generating objections as you go. But when you read a piece of narrative writing that makes the same argument—like Lewis’s profile of Christopher Mark preventing mine collapses—a different process takes over. As you get drawn into the story, your imagination consumes mental resources you’d otherwise use to poke holes in an argument. The story doesn’t silence your inner skeptic: It just keeps that skeptic too busy to speak up.

Stories also have an advantage for Lewis’s goal of shifting public perception. Research finds that when people receive information as a story rather than a statistic, it has a stronger effect on what they believe over time. But this cuts both ways: If your brain holds onto stories more tightly than data, then a vivid anecdote about one incompetent DMV clerk can outweigh a dataset on the efficiency of government work.

This storytelling strategy draws on how other social changes have occurred. Lewis points to the fight for marriage equality as a model. For years, advocates told individual stories and made arguments while facing political resistance. Gradually, as more queer people came out and shared their lives, many Americans came to accept them for who they were. Eventually, this narrative shift created space for political change. Lewis believes civil service needs to follow a similar trajectory: If storytellers show people what’s at stake, public perception of government work can shift enough to create political support protecting essential expertise.

When Narrative Victories Require Ongoing Defense

Lewis’s analogy is instructive, but perhaps not entirely in the way he intends: The issue of LGBTQ+ rights, which he characterizes as an example of a successful narrative change, demonstrates that such victories can erode. Since 2021, overall national support for same-sex marriage has remained around 70%, but this apparent stability obscures a dramatic partisan divergence. A majority of Republicans—55%—supported marriage equality in 2021, but by 2025, that figure had fallen to 41%, erasing nearly a decade of gains. The partisan divide is now 47 points, a record in Gallup’s polling history.

As commentators have noted, once an issue becomes a marker of partisan identity, it absorbs the broader hostility that partisanship generates: People’s feelings about their political opponents start to color their feelings about the issue itself. We can see this dynamic at work in the discourse around LGBTQ+ rights. The marriage equality movement’s core narrative was that same-sex couples are ordinary people living ordinary lives. In conservative spaces, that message has been drowned out by fear-based narratives falsely linking LGBTQ+ people to threats against children, a moral panic that has proven politically powerful precisely because it bypasses reason and triggers protective instincts.

Research on social movements helps explain why some victories hold and others don’t. Experts say the real measure of a movement’s success isn’t whether it achieves its policy goals, but whether the public has come to share the movement’s values by the time those goals are achieved. Policy wins built on broad consensus tend to endure; those achieved without it are vulnerable to backlash. The stories Lewis wants to tell—dedicated workers solving important problems—compete against fear-based narratives (”“lazy bureaucrats waste your money,” “the deep state subverts your will”) that are more attention-grabbing and politically motivating. For Lewis’s project, storytelling may be necessary but not sufficient.

What’s Required From Citizens, Policymakers, and Government?

Lewis’s storytelling strategy requires action from multiple groups. From readers and citizens, Lewis and his collaborators call for shifts in understanding. First, they argue that people should recognize federal workers as among the best in society: people who have figured out how to live meaningful lives by choosing purpose over profit. Second, they contend that citizens should understand government as essential infrastructure rather than overhead to be minimized. Just as society wouldn’t eliminate road maintenance or electrical grids to save money, it shouldn’t eliminate federal workers. Third, they warn against applying the “move fast and break things” mentality, which works in startups, to complex institutions built over decades.

Why Each Shift Asks You to Value What American Culture Doesn’t

Each of Lewis’s requested shifts asks citizens to value the opposite of what American culture typically celebrates. What we’ve been taught to admire—charisma, novelty, or speed—is the enemy of what actually works: humility, maintenance, and institutional patience.

Recognizing federal workers as “the best among us.” Jim Collins’s research in Good to Great found that the leaders who produce the most sustained change aren’t the bold, charismatic people boards typically seek out. Instead, they’re humble but highly driven professionally. They avoid the limelight and tend to credit external forces or colleagues for their successes. Collins calls this “Level 5 leadership,” and he presents it as a data-driven finding because it so thoroughly contradicts our cultural assumption that effectiveness requires self-promotion. Lewis makes an analogous case: The federal workers he profiles are exceptionally talented and driven, but they’ll never make the case for themselves.

Seeing government as infrastructure, not overhead. In The Innovation Delusion, historians Lee Vinsel and Andrew Russell argue that American culture has developed a bias in favor of novelty and disruption. Meanwhile, nobody celebrates the work of those who keep things running. The people who repair, maintain, and steward existing systems tend to be undervalued and underpaid relative to those who build new things. Government workers fit this pattern precisely: Their work is invisible not because it’s unimportant but because it functions quietly behind the scenes, creating the essential systems we take for granted.

Rejecting “move fast and break things.” The idea of “disruption” began as a specific theory about how newcomers unseat older businesses in competitive markets. However, the concept has been stretched far beyond that original context—applied to schools, hospitals, and public institutions as though they were just another industry ripe for reinvention. The problem is that these institutions aren’t serving customers who can switch to a competitor; they serve citizens with complex, ongoing needs. Even Clayton Christensen, who coined the term “disruptive innovation” in The Innovator’s Dilemma, resisted the casual broadening of his framework, arguing that its value depends on applying it precisely.

From policymakers and leaders, Lewis calls for immediate protection and long-term rebuilding. The immediate need is to stop the destruction of the federal workforce that surged in 2025, which could involve advocacy organizations bringing legal challenges, citizens pressuring their representatives, and sympathetic legislators protecting the civil service. Beyond that, Lewis and his coauthors contend that making public service viable for talented young people requires addressing the pay gap, working conditions, and debt burden that prevent them from choosing government careers. They believe society must treat civil service as essential work that deserves respect and adequate compensation rather than as a fallback for those who can’t succeed elsewhere.

The Structural Barriers That Keep People Out of Government Work

Lewis argues that making public service viable requires addressing pay, working conditions, and debt. But the most immediate barrier may be simpler and more fixable: The hiring process drives people away. Two-thirds of Americans aged 18 to 34 agree that federal work could let them make a meaningful difference, yet the same proportion had never even thought about applying. The application process helps explain this gap.

Until recently, federal resumes routinely ran four to six pages, far longer than the private sector’s one-page standard. (The US Office of Personnel Management’s 2025 Merit Hiring Plan has since imposed a two-page cap.) Government job titles often bear little resemblance to the work, some listings close after just a few days, and candidates can wait weeks before learning whether their materials have been reviewed. Then there’s the timeline. In 2024, it took an average of 101 days to bring a new federal employee on board, slow compared to the private sector, where decisions are made within weeks. A new graduate with an offer from a tech company or a consulting firm is unlikely to wait for months to hear from the government.

This problem echoes an argument Alexander Karp and Nicholas Zamiska make about government procurement in The Technological Republic. They describe how procurement rules originally designed to prevent wasteful spending became so cumbersome that the Air Force couldn’t buy necessary electronics during the Gulf War. Federal hiring has followed a similar trajectory: Safeguards built to ensure merit-based, fair selection have calcified into a system so slow that it repels much of the talent it was designed to find. In both cases, well-intentioned protections now prevent the government from getting what it needs.

From the government itself, the authors argue that better communication is necessary. The invisibility that protects workers from political interference becomes a vulnerability when their work is under attack. Lewis contends that agencies need mechanisms to showcase their accomplishments without every statement requiring political approval, and that workers need permission to explain their work to the public. He acknowledges the tension here: The culture of humility he admires also creates the conditions for its own destruction—if nobody knows what you do, nobody defends you when your job’s eliminated.

(Shortform note: Lewis argues agencies need better communication mechanisms, but the obstacles are baked into law. Since 1952, federal law has prohibited agencies from spending funds on “publicity or propaganda” or employing PR staff. Even routine statements can require sign-off from numerous offices. The line between permissible and prohibited communication is fine enough that even big agencies have stumbled, so most stay silent. That silence feeds a vicious cycle: Americans often name dishonesty as their reason for distrusting government messaging, while professionals think agencies need to communicate more clearly. But since agencies are constrained from telling their own stories, someone else has to do it for them.)