THE PADDED SHOVEL
From Madisonian Friction to Administrative Inertia (1776–2026)
A Systemic Analysis of Public-Sector Incentive Decay
The American constitutional order was engineered to harness human ambition through structured friction and visible accountability. Over two centuries, the professional administrative state systematically eliminated that friction, producing a protected class whose dominant output is now institutional self-preservation rather than measurable public service.
PART I: THE PHILOSOPHICAL BEDROCK
The War Against the Parasitic State (1600–1789)
1.1 The Colonial Experience – Extraction Without Consent
To understand why the American Founders built government the way they did, it is necessary to understand the government they had experienced and despised. The British imperial system as practiced in the colonies was not, in their view, a neutral administrative apparatus. It was a machine for extraction — a mechanism by which well-connected men in London converted public authority into private wealth, at the expense of people who had no voice in the matter.
The scaffolding of this system was the patronage network. Royal appointments to colonial offices — governor, customs collector, admiralty judge, surveyor-general — were distributed not on the basis of competence or local knowledge but on the basis of connection, family, and favor. A man who had done useful service for a parliamentary patron, or who had married into the right family, or who had simply outlasted his competitors in the antechambers of Whitehall, might find himself dispatched to Massachusetts or Virginia with a salary, a title, and a set of powers that insulated him almost entirely from the people he nominally served.
The sinecure — an office that paid a salary without requiring meaningful work — was not an aberration of the British system. It was a structural feature. The Court of Exchequer maintained officers whose duties had been obsolete for a century but whose salaries continued uninterrupted. The Post Office distributed lucrative local franchises to political loyalists. Customs houses employed deputy collectors who collected fees without collecting customs. These arrangements were understood, by those who benefited from them and by those who studied them, as the deliberate architecture of a governing class that reproduced itself by controlling access to public income.
For the colonists, the lived experience of this system was the customs officer who boarded ships not to enforce law but to extract bribes; the admiralty court that tried smuggling cases without juries, because juries of local merchants were unlikely to convict; the royal governor who answered to London rather than to the assembly that nominally represented the colony. The writs of assistance — general search warrants that allowed customs officials to enter any premises at will — were not experienced as administrative inconveniences. They were experienced as the physical expression of a government that viewed the governed as a resource to be exploited rather than a constituency to be served.
James Otis argued the writs case in Boston in 1761 with an energy that John Adams, who was present, later said planted the seeds of revolution. His argument was not primarily about search and seizure. It was about the nature of legitimate authority. A government that deployed its powers against the interests of the people it claimed to represent had, in his view, surrendered any claim to obedience. The writs were not just illegal. They were an offense against the basic compact between a people and their government.
This experience — of government as an instrument pointed against the citizen rather than in service of the citizen — was the formative political education of the generation that made the Revolution. It is the context without which the constitutional design cannot be understood. They were not building government from abstract theory. They were building a corrective to something they had lived.
1.2 Thomas Paine and the Language of Parasitic Extraction
Common Sense, published in January 1776, is remembered primarily as a work of revolutionary agitation — the pamphlet that pushed the colonists toward independence. This is accurate but incomplete. Common Sense is also one of the most penetrating analyses of institutional parasitism ever written in plain English, and its economic argument is as relevant to the administrative state of 2026 as it was to the monarchy of 1776.
Paine's foundational distinction is between society and government. Society, he writes, is produced by our wants — by the natural human tendency to associate, cooperate, and produce more together than we can alone. Society is a positive force. Government, by contrast, is produced by our wickedness. It is a restraint that becomes necessary only because people do not always voluntarily honor their obligations to one another. Society is the farmer. Government is the fence. The fence is necessary. But it does not grow the crops.
From this distinction follows a set of inferences that Paine's contemporaries found electrifying and that remain incendiary today. If government is a necessary evil rather than a positive good, then its size, cost, and reach must be proportioned to the minimum necessary for its legitimate functions. Any expansion beyond that minimum is not good governance — it is extraction. It is the fence growing over the field.
Paine is explicit about who benefits from this expansion. The monarchy, the aristocracy, and the established church constitute, in his analysis, a parasitic class that consumes the product of the industrious many without contributing anything equivalent in return. The King's civil list — the annual appropriation for royal household expenses — runs, he notes, to nearly a million pounds per year. This is not compensation for royal service. It is tribute extracted from people who work, paid to people who do not. The hereditary principle ensures that the parasite reproduces itself automatically, without any requirement that the next generation demonstrate fitness for the position it inherits.
The argument has a precision that rewards careful reading. Paine is not making a sentimental case for equality or a metaphysical case for natural rights — though he makes those cases too, elsewhere. Here, his argument is economic. The aristocratic system misallocates talent and resources. Men of ability and energy who lack the right birth are excluded from positions where they could serve the public interest. Men of neither ability nor energy who possess the right birth occupy those positions and produce nothing worth the cost of their occupation. The nation is poorer for this arrangement. The arrangement persists because those who benefit from it control the mechanisms by which it might be reformed.
Read against the backdrop of the modern administrative state, Paine's argument acquires an uncomfortable contemporary resonance. The hereditary principle has been replaced by the civil service examination and the tenure system. The form has changed. The substance — a class of people who extract stable income from the public treasury while remaining substantially insulated from the consequences of their performance — has not changed as much as the reformers who built the modern civil service believed it would.
1.3 The Cincinnatus Ideal vs. Career Officialdom
Lucius Quinctius Cincinnatus was a Roman farmer who, according to legend, was found plowing his fields in 458 BC when messengers arrived to inform him that the Senate had appointed him dictator to deal with a military emergency. He accepted the office, commanded the army, defeated the enemy, disbanded his forces, and resigned the dictatorship — all within fifteen days — then returned to his plow. He had the absolute power of a Roman dictator for two weeks and used none of it for personal advantage. His name became, for centuries afterward, the defining image of what public service was supposed to look like.
George Washington understood himself in these terms and was understood by his contemporaries in these terms. His resignation of his military commission in December 1783 was not merely a formality. It was a deliberate political act, carefully staged, that communicated a specific message to a watching world: in this republic, the military would be subordinate to civilian authority, and civilian authority would be exercised temporarily, not accumulated permanently. King George III, informed that Washington intended to resign and return to private life, is said to have responded that if he actually did so, he would be the greatest man in the world. The comment reflects how genuinely unprecedented the act was in the context of military men with power and popular support.
The Founders' fear of permanent office-holding was not abstract. It was grounded in their reading of history and their understanding of human psychology. Montesquieu, whom they had read carefully, argued that the corruption of republics followed a predictable pattern: citizens who entered public life in a spirit of service gradually came to regard their offices as personal property, to use their positions to enrich themselves and their families, and to resist any accountability that threatened their tenure. The Roman Republic had followed this path. The English Constitution, in their view, was following it in their own time.
The Founders built rotation into the constitutional design wherever they could. The President would serve four-year terms, renewable but not unlimited. Senators would serve six-year terms, staggered so that the entire body could never be replaced at once. Representatives would face the voters every two years — a period short enough that no one could become too comfortable. The point was not merely democratic accountability, though it was that. The point was to prevent the emergence of a permanent governing class whose interest was in perpetuating itself rather than serving the public.
What they could not anticipate, or chose not to confront, was the civil service. The Constitution they designed provided for elected officials and for the officers those officials appointed. It did not — could not — account for the vast permanent bureaucracy that would accumulate over the following two centuries, staffed by people who were neither elected nor easily removed, whose tenure was protected by law, and whose performance was measured by metrics they largely controlled. The Cincinnatus ideal was built into the constitutional architecture. It was not built into the administrative state that grew up around it.
1.4 Anti-Federalist Prophecies: The Warning of the Distant Court
The debate over ratification of the Constitution was, among other things, a debate about institutional distance. The Anti-Federalists — Patrick Henry, George Mason, the pseudonymous Brutus, Centinel, and the Federal Farmer — argued with considerable force that a government operating at the scale the Constitution contemplated would be, by its very nature, unaccountable to ordinary citizens. Their arguments were dismissed at the time as provincial and were largely forgotten after ratification. They deserve to be recovered.
Patrick Henry's speech at the Virginia Ratifying Convention in June 1788 is the most emotionally powerful of the Anti-Federalist arguments, and the least systematic. But buried in Henry's oratory is a precise structural observation: a government with the power to tax at the national level, to maintain a standing army, and to appoint its own officers would inevitably generate a class of officials whose loyalty ran upward to their appointers rather than outward to their constituents. The 'great men,' Henry warned, would hold the power. The people would not know who these men were or how to reach them. They would be, as he put it, 'beyond the Blue Ridge,' operating in a world invisible to those they nominally served.
George Mason's objections were more technical and, in retrospect, more prophetic. Mason had been one of the principal architects of the Virginia Declaration of Rights and was one of the most experienced constitutional draftsmen in America. He refused to sign the Constitution not out of general opposition to federal government but out of specific objections to its structural features. He worried about the lack of a Bill of Rights — he got that remedied. But he also worried about the power of the federal judiciary, whose judges would serve during good behavior — effectively for life — and whose decisions could not be easily overturned. He worried about the treaty power, which could obligate the nation without popular ratification. And he worried about the absence of a council to advise and check the President, which he believed would lead either to a monarch-like executive or to a government dominated by a Senate oligarchy.
Brutus, writing in the New York Journal in 1787 and 1788, produced what many scholars consider the most analytically rigorous of the Anti-Federalist papers. His argument about the federal judiciary has aged particularly well. He predicted, with remarkable accuracy, that a Supreme Court with life tenure, appellate jurisdiction over all federal questions, and no meaningful check on its decisions would accumulate power over time, expanding the scope of federal authority at the expense of the states and the people. The Court would interpret the Constitution in ways that served the interest of federal power, because the Court was itself an instrument of federal power. There was no external check to correct overreach. There was only the self-restraint of the justices — and Brutus was not optimistic about institutional self-restraint.
The Anti-Federalists lost the ratification debate. But the concern that animated them — that distance plus process produces immunity from neighborly judgment — was not refuted. It was simply outvoted. The administrative state that exists today is, in important respects, the fulfillment of their warnings: a permanent class of officials, insulated from elections, operating under rules they largely write and interpret themselves, making decisions that affect millions of people who cannot identify them by name, let alone hold them accountable.
1.5 Madison's Realistic Engineering: Ambition Counteracting Ambition
James Madison was not a utopian. He did not believe that human nature could be reformed by education or moral instruction to the point where good government could be trusted to the virtue of its officials. He believed, on the contrary, that human beings were reliably self-interested, that factions — groups of citizens united by a common interest adverse to the interests of other citizens or the community as a whole — were inevitable in any free society, and that the task of constitutional design was not to eliminate this reality but to work with it.
Federalist No. 10, published in November 1787, is the locus classicus of this argument. The problem of faction is, Madison says, the mortal disease of popular governments throughout history. Pure democracies — small societies governing themselves directly — have always been convulsed by faction and have always collapsed into either tyranny or chaos. The solution is not to suppress faction — that would require destroying the liberty that makes faction possible — but to extend the sphere of the republic to the point where no single faction can dominate. A larger republic contains more factions, and the competition among them makes it harder for any one group to seize control of the whole.
Federalist No. 51 extends the argument to the internal structure of government. The legislative, executive, and judicial branches must each have sufficient independence and sufficient power to resist encroachments by the others. Each officer must be given 'the necessary constitutional means, and personal motives, to resist encroachments of the others.' The key phrase is personal motives. Madison is not relying on a sense of constitutional duty or institutional loyalty. He is relying on self-interest. If the ambitions of the men in each branch are pointed at the ambitions of the men in the other branches, the friction produced by their competition will protect the liberty of the citizen.
'Ambition must be made to counteract ambition,' Madison writes. This is the machine. It does not require virtue to run. It runs on rivalry, jealousy, and the desire for power. The President checks the Congress because the President does not want Congress to dominate. Congress checks the President because Congress does not want to be dominated. The judiciary checks both because the judiciary has an institutional interest in its own prerogatives. No branch is virtuous. All are ambitious. The competition among their ambitions produces, as a byproduct, something that looks like accountability.
What Madison could not have designed for was a situation in which the competition among branches was replaced by a competition between elected officials and a permanent civil service — a competition that the elected officials were structurally positioned to lose. The civil service has continuity, expertise, and institutional memory that no elected official can match. It outlasts any administration. It controls the information that elected officials need to make decisions. It can, through delay, obstruction, and selective disclosure, substantially constrain the effective authority of officials who nominally govern it. Madison's machine works when the competing parties have roughly equivalent resources. It misfires when one party has overwhelming advantages in information and tenure.
1.6 The ‘Hole’ Principle in the Early Republic
The early American republic was, by any modern standard, almost inconceivably small. The federal government in 1790 employed approximately 350 people, not counting the military. The entire executive apparatus — the State Department, the Treasury, the War Department, the Attorney General's office — could have fit comfortably in a single building, had a suitable building existed. The federal budget was measured in millions of dollars rather than trillions. The government's reach into daily life was, for most Americans, negligible.
This smallness was not merely a function of limited resources. It was, in significant part, deliberate. The Founders were not trying to build an administrative state. They were trying to build a minimal government that could perform the functions necessary for national security and commercial order — defense, foreign policy, currency, the postal road, the enforcement of contracts — without acquiring the capacity to oppress its citizens. The 'hole' in the shovel was not padded because there was almost no shovel.
Local government was a different matter. Towns governed themselves through direct democratic institutions — the town meeting, the board of selectmen, the jury — in which accountability was immediate, visible, and personal. The tax collector lived in the same town as the people he taxed. The road commissioner could be confronted at church about the road. The school committee member's children attended the school. The performance of public officials was evaluated not by metrics designed by those officials but by the judgment of neighbors who experienced the results directly.
Alexis de Tocqueville, visiting the United States in 1831, was struck by this feature of American public life more than almost any other. In France, he observed, the central government performed an enormous range of functions — poorly, slowly, at great expense, and with little accountability to the people it served. In America, the same functions were performed locally, by citizens who rotated in and out of office, who were directly accountable to their neighbors, and who had no professional stake in the perpetuation of their positions. The result, in Tocqueville's assessment, was government that was sometimes rough and amateurish but was rarely corrupt in the particular way that professional bureaucracies tend toward corruption — the corruption of insulation, of captured expertise, of the official who has learned to serve the institution rather than the public.
This was the hole before it was padded. Tangible output was the only acceptable metric because there was no bureaucratic structure to substitute process for results. The road was either passable or it was not. The bridge either held or it collapsed. The tax collector either produced revenue or he was replaced. The feedback loop between official action and citizen judgment was short, direct, and difficult to manipulate.
PART II: THE GREAT INVERSION
Professionalism as Insulation (1800–1980)
2.1 Hamilton vs. Jefferson: The Original Tension Over Scale
The first great constitutional conflict of the early republic was, at its core, a conflict about the nature of the administrative state before the administrative state existed. Alexander Hamilton wanted a government capable of performing complex functions efficiently — managing a national debt, chartering a national bank, promoting industry, projecting power abroad. Thomas Jefferson wanted a government that remained close to the people, visible and accountable, operating at a scale that ordinary citizens could comprehend and influence.
Hamilton's vision required something Jefferson's did not: a competent professional class of administrators. You cannot manage a national bank with town-meeting democracy. You cannot conduct foreign policy by popular referendum. Expertise, continuity, and institutional memory are genuine requirements for some governmental functions, and Hamilton understood this with a clarity that Jefferson, absorbed in his agrarian ideal, sometimes seemed to resist.
Jefferson's vision required something Hamilton's did not: simplicity. A government that ordinary farmers could understand was a government that ordinary farmers could hold accountable. A government that required specialists to comprehend was a government that only specialists could evaluate. The moment government becomes too complex for the citizen to audit, the citizen loses the practical capacity to govern, regardless of what the formal institutions say.
Both men were right about something. Hamilton was right that some governmental functions genuinely require professional expertise and that the pretense of universal competence is its own form of dysfunction. Jefferson was right that complexity is the enemy of accountability and that a professional class with a monopoly on governmental knowledge will, over time, serve its own interests as much as the public interest.
The constitutional settlement was ambiguous in ways that allowed both visions to persist in tension. The federal government was given the powers Hamilton thought necessary. The Bill of Rights and the system of reserved powers gave Jefferson's concerns their formal expression. What neither man fully anticipated was the dynamic by which professional expertise would accumulate faster than democratic accountability could adapt — the dynamic by which the Hamiltonian administrative apparatus would become, over two centuries, the kind of entrenched governing class that Jefferson had feared and that the colonial experience had taught a generation to despise.
2.2 The Spoils System and Its Crisis
Andrew Jackson's election in 1828 was, among other things, a revolt against the idea of a permanent governing class. The men who had run the federal government since its founding were, in Jackson's telling, a self-perpetuating elite whose claim to office rested on education, family connection, and mutual back-scratching rather than democratic mandate or demonstrated service to ordinary Americans. His solution was rotation in office: the systematic replacement of incumbent officeholders with loyal supporters.
The spoils system — 'to the victor belong the spoils,' as Senator William Marcy of New York memorably put it — had a democratic logic. If offices belong to the people rather than to the men who hold them, then the people's elected representatives should be able to staff the government with their own people. This is not obviously wrong. The alternative — a permanent civil service whose members cannot be removed regardless of which party wins elections — creates its own accountability problems, as we have since discovered.
But the spoils system had catastrophic practical consequences. Offices were distributed to men chosen for their political loyalty rather than their competence. The quality of public administration deteriorated. Corruption was institutionalized: office-holders who knew their tenure depended on the next election result had every incentive to use their offices to enrich themselves before the music stopped. The post offices, the customs houses, the land offices — all became machines for extracting value from the public while providing minimal service in return.
The crisis came in July 1881, when President James Garfield was shot in a Washington train station by Charles Guiteau, a disappointed office-seeker who had applied for a consulship and been rejected. Guiteau's grievance was not exotic. He had done political work for the Republican Party, he believed he was owed an appointment, and he had spent months haunting the State Department and the White House trying to collect what he considered his due. When the appointment did not come, he decided that Garfield had to die so that Chester Arthur — whom Guiteau imagined to be more sympathetic to the Stalwart faction of the party — could become President and correct the injustice.
Garfield died in September. The political consequences were immediate. The assassination had made the spoils system literally lethal. It had produced a class of people who regarded public offices as personal entitlements, who experienced the denial of those entitlements as an outrage requiring violent redress, and who had been encouraged in this view by decades of political culture that treated offices as rewards for partisan service. The reformers who had been arguing for a merit-based civil service for years suddenly had the argument they needed. The Pendleton Civil Service Reform Act became law in 1883.
2.3 The Pendleton Act of 1883 – Reform with Unintended Consequences
The Pendleton Act was, by the standards of its time, a genuine reform. It created a Civil Service Commission charged with administering competitive examinations for a defined class of federal positions. Appointments to those positions would be based on examination scores rather than political connections. The 'classified service' — the positions covered by the Act — initially comprised about twelve percent of the federal workforce. Over the following decades, as successive Presidents extended civil service protections by executive order, the proportion grew until the great majority of federal employees were covered.
The intended effect was clear: to replace a system based on patronage and political loyalty with a system based on demonstrated competence. The unintended effect was equally clear, though it took longer to manifest: to create a class of employees whose job security was essentially permanent, who could not be removed for poor performance without a protracted procedural ordeal, and who therefore had no strong incentive to perform well.
The logic of the transition is important to understand. Under the spoils system, the problem was that employees could be removed too easily — for purely political reasons, without regard to their performance. The solution was to make removal more difficult. But making removal more difficult for political reasons also made removal more difficult for performance reasons. The law did not distinguish between a good employee being harassed out of her job by a politically hostile supervisor and a mediocre employee being appropriately disciplined for poor work. It treated both as equally protected.
The practical result, over time, was the institutionalization of job security as the dominant characteristic of public employment. Private-sector employees knew that their continued employment depended, to a meaningful degree, on the quality of their work. Public-sector employees in the classified service knew that their continued employment depended, primarily, on avoiding explicit violations of clearly articulated rules. The incentive structure had shifted from 'perform well to keep your job' to 'don't get fired.' These are not the same thing.
The seniority norm accelerated this dynamic. In most public-sector systems, seniority — time in service — became the primary basis for advancement, for assignment of desirable duties, and for protection against layoff. A senior employee who had stopped trying to do good work ten years ago was more protected than a junior employee who was doing excellent work. The signal this sent to rational observers of the system was unambiguous: longevity is what matters. Results are secondary.
This is the birth of the padded shovel. The shovel still existed. The work still nominally got done. But the cushion between effort and consequence had been installed. The man digging the hole could now lean on the handle without losing the shovel.
2.4 The Judicial Padding: Qualified Immunity (1967 onward)
The doctrine of qualified immunity — the legal shield that protects government officials from civil liability for constitutional violations unless they violated 'clearly established' law — was invented by the Supreme Court in 1967 and has been expanding ever since. Its history is a case study in how courts create law under the guise of interpreting it, and in how a narrow exception designed to address a genuine problem can be extended into a sweeping protection that effectively immunizes an entire class of government conduct.
The immediate context was the Civil Rights Act of 1871, codified at 42 U.S.C. §1983, which created a federal cause of action against state officials who violated constitutional rights. Congress passed the statute during Reconstruction specifically to address the systematic failure of Southern state courts to hold officials accountable for violence against freed people and their allies. The statute's language is unambiguous: 'Every person who, under color of any statute... subjects, or causes to be subjected, any citizen of the United States... to the deprivation of any rights, privileges, or immunities secured by the Constitution and laws, shall be liable to the party injured in an action at law.'
Every person. No exceptions for good faith. No exceptions for uncertainty about the law. The text does not say 'every person unless they reasonably believed their conduct was lawful' or 'every person unless the constitutional rule they violated had been clearly established by prior precedent.'
Pierson v. Ray, decided in 1967, introduced the good-faith defense for the first time. The case involved police officers who had arrested civil rights demonstrators for breach of the peace under a Mississippi statute that was later held unconstitutional. The Court held that the officers could not be held liable because they had acted in good faith and with probable cause. This was a defensible holding on its particular facts — holding officers liable for enforcing a statute that had not yet been ruled unconstitutional seemed harsh. But the reasoning contained the seed of something much larger.
Harlow v. Fitzgerald, decided in 1982, planted that seed more deeply. The Court held that officials were entitled to immunity unless they violated 'clearly established statutory or constitutional rights of which a reasonable person would have known.' The phrase 'clearly established' became, in subsequent decades, a term of art with a very specific and very demanding meaning. A right is not 'clearly established' simply because it is stated in the Constitution or in a Supreme Court decision. It is 'clearly established' only if there is prior precedent, in the same jurisdiction, involving sufficiently similar facts, that would put a reasonable officer on notice that the specific conduct at issue was unconstitutional.
The practical effect of this standard is that officers can often violate constitutional rights without liability simply because no prior case has addressed the precise factual combination present in their case. The doctrine has been criticized by legal scholars across the ideological spectrum, including by some of the Justices themselves. Justice Clarence Thomas, in a 2021 dissent, noted that the doctrine as it has developed has no basis in the text of the statute, no support in the common-law tradition at the time of the statute's passage, and no clear justification in precedent. It was, he suggested, an exercise in judicial legislation that Congress should address.
Whatever its doctrinal deficiencies, qualified immunity has had a concrete and measurable effect on the accountability of government officials. It has made it substantially harder for citizens to obtain redress for constitutional violations. It has insulated officials from the financial consequences of misconduct. It has, in combination with civil service protections, contributed to an environment in which government officials face lower accountability for their actions than private individuals face for equivalent conduct.
2.5 The Administrative State as a De Facto Fourth Branch
The Constitution establishes three branches of government. The administrative state that has developed over the past century and a half constitutes, in functional terms, a fourth branch — one that combines legislative, executive, and judicial powers in a single set of institutions, and that is insulated from democratic accountability in ways that none of the three formal branches are.
Administrative agencies write rules that have the force of law — a legislative function. They enforce those rules — an executive function. They adjudicate violations — a judicial function. And they do all of this with a permanence and continuity that no elected official can match. An elected president serves, at most, eight years. The career officials of the Environmental Protection Agency, the Securities and Exchange Commission, the Department of Health and Human Services, and the hundreds of other federal agencies serve indefinitely. They outlast the political appointees who nominally direct them. They retain the institutional knowledge, the professional networks, and the practical expertise that enable them to operate with substantial independence from whoever happens to be in elected office.
The Administrative Procedure Act of 1946 created a framework for agency rulemaking and adjudication that was designed to ensure procedural fairness and judicial review. It largely succeeded at both. But procedural fairness and judicial review are not the same as democratic accountability. An agency that follows all the required procedures for issuing a rule — publishing the proposed rule, soliciting comments, responding to significant comments, publishing the final rule — has complied with the APA regardless of whether the rule serves the public interest. The process is the output. Compliance with procedure is what is measured. The actual effects of the rule on actual people are evaluated, if at all, by mechanisms that are slow, expensive, and uncertain.
The Chevron doctrine — established by the Supreme Court in 1984 and substantially overruled in Loper Bright Enterprises v. Raimondo in 2024 — added another layer of insulation. Under Chevron, courts were required to defer to agencies' interpretations of ambiguous statutory language as long as those interpretations were 'reasonable.' This meant that agencies could, within limits, determine the scope of their own authority. The referees were also players. The recent overruling of Chevron represents a significant correction, but the practical culture of deference it created in courts and among regulated parties does not disappear overnight because a doctrine has been abandoned.
The result of these developments is an administrative apparatus that is, in important respects, more powerful than the formal constitutional branches. It makes more law, through rulemaking, than Congress does through legislation. It resolves more disputes, through administrative adjudication, than the federal courts do through litigation. It operates with a continuity and a depth of expertise that no elected body can replicate. And it is accountable to the public primarily through elections of officials who appoint its leadership — a chain of accountability long enough and indirect enough that its practical effect on day-to-day administrative behavior is limited.
PART III: THE SEVEN LAYERS OF INSTITUTIONAL INERTIA
The Analytical Core
The system of incentives that governs the modern public sector does not produce corruption in the obvious sense — bribery, embezzlement, deliberate malfeasance — as its primary output. It produces something subtler and, in aggregate, more costly: a systematic preference for institutional self-preservation over public service, for process compliance over measurable results, and for the protection of the individual official over the welfare of the citizen who depends on that official's competence and honest effort.
This preference is not the product of bad character. Most public-sector employees are neither corrupt nor negligent in any morally culpable sense. They are, like most people, rational actors responding to the incentive structures they actually face rather than the ones they are supposed to face. The seven layers described below are the structural features of modern public administration that produce the padded-shovel outcome: a system that goes through the motions of public service while generating, as its primary output, the perpetuation of the system itself.
Layer 1: Economic Decoupling – The Stable Salary
The most fundamental structural feature of the modern public sector is the near-complete decoupling of individual compensation from individual performance. In a competitive market, an employee who produces significantly less than her compensation costs will eventually lose her job. In a well-functioning organization, an employee who performs exceptionally well will be compensated more, promoted faster, or recruited by competitors who value her contribution. The relationship between performance and reward is imperfect in the private sector, but it exists as a meaningful force.
In the public sector, the force is weak to nonexistent. The federal General Schedule pay system, which covers approximately 1.5 million civilian employees, determines compensation primarily by grade and step — two variables that reflect, respectively, the classification of the position and the number of years the employee has held it. Neither variable measures what the employee actually accomplishes. A GS-12 Step 7 employee who has spent seven years doing mediocre work receives the same salary as a GS-12 Step 7 employee who has spent seven years doing excellent work. The system contains no mechanism for distinguishing them.
Annual performance ratings exist on paper. In practice, the distribution of ratings is heavily compressed at the upper end. An analysis of federal performance ratings from 2019 found that approximately 71 percent of federal employees received ratings of 'Exceeds Fully Successful' or 'Outstanding' — the two highest categories. Less than one-tenth of one percent received ratings of 'Unacceptable,' the lowest category. In a workforce of 1.5 million, a normally distributed performance curve would produce far more variation. What this distribution actually reflects is not a remarkably high-performing workforce. It is a system in which supervisors have learned that issuing low ratings creates procedural burdens and interpersonal conflict that exceed any benefit from accurate evaluation.
The pension and benefit structures of public employment add another layer of economic insulation. Defined-benefit pensions, available to most state and many federal employees, guarantee a fixed income in retirement based on years of service and final salary rather than on any measure of career contribution. Health benefits typically exceed those available in private employment. Job security, as discussed in the Pendleton Act section, is substantially greater. The result is a compensation package that, for most public employees, is meaningfully superior to what they could command in private employment for equivalent work, and that is almost entirely insulated from performance.
The rational response to this incentive structure is not laziness in any simple sense. It is a particular kind of strategic calibration: invest effort sufficient to avoid explicit discipline, and invest the remainder in the activities most likely to produce advancement — which, in a system where performance is weakly evaluated, tend to be activities that generate favorable notice from supervisors (relationship management, visible compliance with directives, participation in institutional rituals) rather than activities that produce measurable benefits for the public.
Layer 2: Fragmentation of Memory – The Fresh Start
Perhaps the single most consequential structural feature of the modern administrative and judicial state — more consequential, even, than tenure protection or qualified immunity — is its systematic refusal to maintain and act upon longitudinal knowledge about the actors it encounters repeatedly. Every interaction is treated, procedurally if not always practically, as if it were the first.
This feature is partly intentional and partly structural. It is intentional in the sense that due process principles require that each proceeding be evaluated on its own merits, that prior adverse findings not automatically predetermine current outcomes, and that individuals not be punished for past conduct in proceedings addressing present conduct. These are genuine values, important to the integrity of any adjudicative system.
But the fresh-start norm has metastasized far beyond its principled core. Law enforcement agencies that process complaints from the same individuals repeatedly do not maintain the analytical infrastructure to identify patterns across reports. Prosecutors who encounter the same actors through multiple proceedings do not have a systematic mechanism for evaluating patterns of conduct rather than isolated incidents. Judges who see the same parties more than once are formally constrained in what prior information they can consider. Courts in different counties and states have no mechanism to share the accumulated record of a repeat actor's behavior across jurisdictions.
The practical consequence is that a sophisticated actor who understands the fresh-start norm can exploit it systematically. By generating a sufficient volume of interactions — each individually below the threshold of action, each involving different officials in different proceedings, each presenting a narrative that has not yet been contradicted by the full accumulated record — an actor can impose enormous costs on a target while each individual institutional response is technically correct in isolation.
The protection this structure provides to institutional actors rather than individual ones is equally significant. An agency that has failed to perform its functions adequately in a series of discrete cases will rarely be confronted with an accounting of its pattern of failure. Each case is closed individually. The aggregate is nobody's responsibility. The official who signed the inadequate report in Case A is not the official who closed Case B without investigation and is not the official who processed Case C without contacting the affected party. Nobody owns the pattern. Nobody is accountable for the cumulative outcome.
The fresh start, extended to institutions, functions as a structural amnesia that makes learning from failure essentially impossible. Organizations that cannot acknowledge their patterns of failure cannot correct them. The padded shovel remains padded because nobody is measuring the depth of the holes.
Layer 3: Risk Aversion and Easy Metrics
Government officials are evaluated, to the extent they are evaluated at all, on metrics that measure activity rather than outcomes. The police department's performance metrics track arrests made, calls responded to, and response times recorded. They do not track whether the arrests led to convictions based on solid evidence, whether the calls were handled in ways that actually resolved the underlying problems, or whether the response times reflected appropriate prioritization of serious matters over trivial ones. The number of reports filed is easy to count. Whether the reports accomplished anything is harder to measure and riskier to evaluate.
This produces a well-documented phenomenon in organizational behavior: when metrics substitute for outcomes, organizations optimize for the metrics rather than for the outcomes the metrics were meant to represent. Police departments that measure arrests will maximize arrests, including arrests that do not serve public safety. Prosecutors who measure conviction rates will pursue easy cases and decline difficult ones, even when the difficult cases involve more serious harm. Social service agencies that measure caseload closures will close cases, regardless of whether the underlying problems have been addressed.
The risk-aversion dynamic compounds this problem. Government officials face asymmetric consequences: taking an action that turns out to be wrong creates visible accountability — a complaint, a grievance, a lawsuit, a media story. Failing to take an action that would have been right creates no comparable consequence, because the counterfactual cannot be observed. The detective who charges the wrong person faces scrutiny. The detective who declines to investigate at all faces nothing. The prosecutor who brings a weak case faces adverse publicity if it fails. The prosecutor who declines to bring a strong case faces nothing.
The rational official, navigating this asymmetry, will systematically underprovide action on complex, contested, or politically sensitive matters while systematically overproviding action on simple, clear-cut matters where the probability of error is low. This means that the people who most need government intervention — those whose situations are complex, whose adversaries are sophisticated, and whose claims are contested — are systematically underserved relative to those whose needs are simple, whose adversaries are weak, and whose claims are straightforward.
The triage logic of prosecutorial discretion illustrates this pattern clearly. Prosecutors operating under caseload pressure will rationally allocate their effort to cases that are most likely to produce convictions with least expenditure of resources. These are typically cases with clear evidence, cooperative victims, and relatively unsophisticated defendants. Cases that involve complex patterns of conduct spread across multiple incidents, sophisticated defendants with resources to mount a defense, or victims whose credibility is contestable are disproportionately declined. The formal justification is resource constraints. The structural reality is that the metrics by which prosecutors are evaluated do not reward the difficult case.
Layer 4: Institutional Loyalty – The Blue Wall and Its Civilian Equivalents
The 'blue wall of silence' — the norm within law enforcement cultures against reporting or cooperating with investigations of fellow officers — is the most visible and most discussed version of a phenomenon that exists throughout the public sector. It is not unique to law enforcement. It is not unique to the public sector. But it takes a particularly consequential form in institutions that hold coercive power over citizens and that face limited external oversight.
The norm is not primarily maintained by explicit threats, though explicit threats sometimes occur. It is maintained by the social and professional structure of the institution. Officers work together in high-stress situations where they depend on one another for safety. The trust that makes this possible requires a degree of unconditional loyalty that extends beyond the job itself. To report a colleague's misconduct is to betray that trust, to be labeled a 'rat,' to be excluded from the informal social network that makes working conditions tolerable, and to face the practical reality that colleagues who distrust you are colleagues who may not have your back when it matters.
This structure produces a systematic bias in internal investigations, in supervisory evaluations, and in the institutional response to misconduct complaints. Supervisors who investigate complaints against their subordinates are investigating people whose cooperation they depend on. Internal affairs units are staffed by officers whose career advancement depends on the goodwill of the departments they investigate. The institutional incentive, in nearly all cases, is to find insufficient evidence of serious misconduct, to attribute problematic behavior to stress or misunderstanding, and to close the matter with minimal disruption to institutional operations.
The civilian equivalents of the blue wall operate on the same logic with less dramatic consequences in individual cases but similar aggregate effects. The social worker who suspects a colleague of falsifying case records faces the same basic calculus as the officer who witnesses misconduct: reporting is socially costly, professionally risky, and unlikely to produce a proportionate benefit. The benefits of reporting accrue primarily to the public, and specifically to members of the public who are often invisible to the institution. The costs accrue personally and immediately to the person who reports.
The institutional response to whistleblowing across the public sector is consistent enough to constitute a pattern. Formal whistleblower protections exist in law. In practice, the person who reports institutional misconduct faces: informal ostracism, loss of desirable assignments, adverse performance evaluations, disciplinary proceedings triggered by suddenly-discovered violations of rules that were previously ignored, and the slow erosion of career prospects that comes with being known as someone who causes trouble. The formal protection against retaliation does not prevent the informal mechanisms of institutional punishment. It creates a paper record that the institution has followed the rules while the realities of workplace life make the whistleblower's position untenable.
Layer 5: Procedural Formalism – The Report Is the Result
Modern public administration has developed an elaborate set of procedures for documenting, processing, and formally resolving the matters that come before it. These procedures serve genuine purposes: they create records, constrain arbitrariness, and provide the basis for review. But they have also generated a professional culture in which the completion of procedure has become, functionally, the definition of performance.
The report is the result. This is not a cynical formulation — it is an accurate description of how most public institutions evaluate their own performance. A complaint was received. A report was filed. A determination was made. The box was checked. Whether the determination was correct, whether the process produced an outcome that served the interests of the person who filed the complaint, whether the problem that generated the complaint was actually addressed — these questions are not absent from the formal evaluation framework, but they are substantially harder to measure than whether the report was filed on time and in the correct format.
This shift from outcomes to outputs has been documented extensively in the public administration literature under the heading of 'goal displacement' — the substitution of means for ends. The means — the procedures, the reports, the forms, the timelines — were adopted in the first place because they were believed to correlate with good outcomes. Over time, the correlation weakens as institutions learn to optimize for the means without attending to the ends. The procedure is followed meticulously. The outcome does not occur. The institution nevertheless records a performance.
In the judicial system, this dynamic manifests in the processing of cases without substantive engagement with the evidence. Hearings are held within the required timeframe. Decisions are issued in the required format. The formal requirements are met. Whether the decision reflected careful analysis of the available evidence is not reliably evaluated, because the formal record contains only the hearing and the decision — not the quality of the reasoning that connected them.
In law enforcement, it manifests in the processing of complaints without investigation. A complaint is received. An officer is dispatched. A report is filed. The matter is closed. Whether the report was based on interviews with all relevant parties, whether available video evidence was reviewed, whether the complaint was part of a pattern requiring a different response — none of these questions are answered by the fact that a report was filed and the matter closed. But a closed report is what the metrics record.
The practical effect is a system that generates large volumes of procedural output — reports, hearings, determinations, closures — while producing a much smaller volume of actual service. The citizen who needs help is likely to receive procedure. The citizen who needs procedure is likely to receive a thorough response. These are not the same populations.
Layer 6: Somatic and Cultural Normalization
Institutions that have operated below their formal performance standards for long enough develop a cultural infrastructure that normalizes the gap between what they claim to do and what they actually do. This normalization is not a conspiracy. It does not require anyone to consciously decide that standards should be lower. It is the natural result of cognitive dissonance resolution in a stable institutional environment.
The officer who has worked in a department where complaints are routinely processed without investigation for fifteen years has adapted to that reality. It is, for him, simply how things work. The prosecutor who has observed for a career that complex pattern cases are routinely declined has internalized that as the appropriate threshold for action. The judge who has seen the same type of claim dismissed for a decade without appellate reversal has developed a default expectation about what those claims are worth. None of these individuals need to be corrupt, or lazy, or malicious. They have simply been shaped by an environment that provides continuous feedback that their current practice is adequate.
The cultural normalization has physical manifestations. Energy levels in meetings about institutional performance tend to track the degree to which participants believe the performance standards are meaningful. In institutions where the gap between formal standards and actual practice is wide and acknowledged only obliquely, the energy for genuine improvement is low. The conversations about performance tend to focus on the metrics that are being met rather than the outcomes that are not being achieved. The implicit message is that the metrics are real and the outcomes are aspirational.
New employees are socialized into this culture through the ordinary mechanisms of professional development. They are supervised by people who have adapted to the existing standards. They are evaluated by criteria that reflect the existing culture. They are rewarded for exhibiting the behaviors that their supervisors exhibit. The institutional culture reproduces itself, generation by generation, in people who never consciously choose to perpetuate it.
The organizational literature describes this as 'institutional isomorphism' — the tendency of organizations facing similar pressures to become structurally similar regardless of whether that structure serves their nominal purposes. Public agencies in the same legal environment, subject to the same oversight mechanisms, staffed by professionals trained in the same schools, develop similar cultures of procedural compliance and outcome indifference not because anyone designed them that way but because the pressures they face systematically reward that culture.
Layer 7: Capture of the Expert Class
The most intellectually interesting layer of institutional inertia is the one that involves the most intellectually sophisticated participants: the credentialed professionals whose expert knowledge provides the institutional legitimacy that allows the system to continue. Lawyers who understand clearly that a particular legal argument is circular but advance it anyway because their client is the institution. Policy analysts who recognize that a program is not working but produce evaluations that emphasize its procedural compliance. Investigators who know that a pattern exists but write reports that treat each incident in isolation because that is the institutional norm.
These are not stupid people. They are, in many cases, the smartest people in their institutions. What has happened to them is not a failure of intelligence. It is a failure of the mechanisms that are supposed to translate intelligence into honest judgment.
The career incentive structure in professional public service rewards compliance over conscience. The lawyer who advances the circuitous argument gets paid regardless of whether the argument is persuasive. The policy analyst who produces a favorable evaluation gets promoted if her supervisors value favorable evaluations. The investigator who adopts the institutional norm of incident isolation faces no professional consequence for doing so. In each case, the intelligent professional has been placed in an environment where intelligence is deployed in the service of institutional preservation rather than in the service of the people the institution nominally serves.
The phenomenon has been studied in the context of financial regulation under the concept of 'regulatory capture' — the tendency of regulatory agencies to be dominated over time by the industries they regulate. But capture is a broader phenomenon than its regulatory applications suggest. Any institution that relies on expert knowledge, that evaluates the quality of expert work primarily through internal processes, and that provides strong career incentives for conformity to institutional culture will tend to capture the intelligence of its most capable participants in the service of its own interests.
The defense that smart people in public institutions often offer for their participation in obviously suboptimal systems is that the alternative — resignation, public criticism, whistleblowing — would accomplish nothing while destroying their careers. This is often true. The individual's rational calculation is correct. The aggregate effect of all the individually rational calculations is a system that continues to function badly, protected from reform by the very intelligence of the people who know it best.
PART IV: THE NATIONAL SCALE
The Trough Expanded
4.1 GAO Improper Payments Reports – The Aggregate Toll
Since fiscal year 2003, when the Improper Payments Information Act required federal agencies to identify and report payments made in error — payments made in the wrong amount, to ineligible recipients, for goods or services not received, or through other violations of applicable law or regulation — the cumulative total of such payments has exceeded $2.8 trillion. In fiscal year 2023 alone, the reported total was approximately $236 billion.
These numbers require some unpacking. Not all improper payments represent fraud or deliberate misconduct. A substantial portion are administrative errors: overpayments due to incorrect data, underpayments corrected only after the fact, payments processed against outdated eligibility information. Some improper payments are recovered after detection. The headline numbers overstate the net loss to the public fisc in some respects and understate it in others — not all improper payments are detected, and the known figures represent only what agencies are required to report under current law.
But the aggregate figure, properly understood, reflects something more important than its accounting implications. It reflects the practical consequence of the structural features described in Part III: a system in which the metrics rewarded are process compliance and volume rather than payment accuracy; in which the institutional cost of tightening verification exceeds the institutional benefit, as defined by the metrics that actually govern behavior; and in which the people who bear the cost of improper payments — taxpayers who funded the overpayments, legitimate beneficiaries whose underpayments reflect the same systems failures — have no direct mechanism to impose accountability on the officials responsible.
The Medicare and Medicaid programs have consistently accounted for the largest share of improper payments in recent years — a combined total that has exceeded $100 billion in some fiscal years. The persistence of these figures despite years of reform efforts illustrates the limits of internal reform mechanisms in systems where the structural incentives for accurate payment processing are weak. The Centers for Medicare and Medicaid Services is aware of its improper payment problem. It has implemented numerous corrective action plans over the past twenty years. The problem persists because the corrective action plans address symptoms rather than the structural features that produce the symptoms.
4.2 Pandemic Unemployment Insurance Fraud
The COVID-19 pandemic produced one of the largest and most rapidly executed expansions of public benefit programs in American history. The CARES Act, enacted in March 2020, extended unemployment insurance to categories of workers not previously covered — gig workers, self-employed individuals, independent contractors — through the Pandemic Unemployment Assistance program, and substantially increased benefit amounts through the Federal Pandemic Unemployment Compensation supplement. At the peak, the federal government was distributing approximately $50 billion per week in unemployment benefits.
The scale and speed of the expansion created conditions in which fraud flourished. Normal eligibility verification procedures were suspended or attenuated in the name of getting benefits to people quickly. States, overwhelmed by claim volumes, processed applications without the identity verification, employer confirmation, and earnings documentation that the ordinary system required. Criminal enterprises — some domestic, some international — obtained personal identifying information through data breaches and other means and filed claims on behalf of people who had not filed, people who had not lost their jobs, and people who did not exist.
The Government Accountability Office estimated in 2023 that fraud and improper payments in pandemic unemployment programs may have totaled as much as $135 billion, with some estimates running higher. The true figure may never be known because state agencies' record-keeping was insufficient to support precise reconstruction of the full universe of fraudulent claims.
The pandemic fraud episode is often framed as an exceptional case — an unprecedented emergency that overwhelmed systems designed for normal conditions. This framing contains truth but is also self-serving for the institutions involved. The identity verification systems that were suspended had been known to be inadequate before the pandemic. The agency systems for detecting patterns of fraudulent applications were known to be weak. The accountability mechanisms for state unemployment offices — which administered the federal programs — were known to be indirect and slow. The pandemic did not create these weaknesses. It revealed them at scale.
More instructively, the fraud that occurred during the pandemic was substantially perpetrated by actors who understood how to exploit the fresh-start norm — filing claims in multiple states simultaneously, using stolen identities that would not appear on any single agency's fraud watchlist, generating a volume of activity that overwhelmed the analytical capacity of systems designed to process claims rather than to detect patterns. The structural vulnerability was not new. The pressure that exposed it was.
4.3 Minnesota's Feeding Our Future Case
The federal Child Nutrition Programs — the National School Lunch Program, the School Breakfast Program, and the Summer Food Service Program — reimburse organizations that provide meals to low-income children. The programs are administered by the United States Department of Agriculture, which distributes funds to state agencies, which in turn reimburse participating organizations based on the number of meals served to eligible children.
In Minnesota between 2020 and 2022, an organization called Feeding Our Future and its associated network of approximately 70 meal sites defrauded the program of more than $250 million. The mechanism was straightforward: the sites claimed reimbursement for meals that were never served to children who did not exist. At some sites, the claimed number of meals exceeded the total population of children in the area by factors of ten or more. Money that was supposed to feed children flowed instead into real estate purchases, luxury vehicles, wire transfers to foreign accounts, and personal enrichment of the scheme's participants.
The fraud was eventually detected and prosecuted. By 2024, dozens of defendants had been charged, and many had been convicted. But the more revealing story is how long the fraud operated before detection and why.
The Minnesota Department of Education, which was responsible for overseeing the program at the state level, received numerous complaints and red flags during the period of the fraud. Whistleblowers identified specific sites where claimed meal counts were implausible. Auditors flagged concerns about documentation. Oversight officials raised questions about the expansion of the program and the implausible speed at which new sites were being approved.
The institutional response to these concerns was shaped by each of the seven layers of inertia identified in Part III. Economic decoupling: the officials responsible for oversight faced no personal financial consequence from approving fraudulent claims. Fresh-start norm: each application was processed individually rather than as part of a pattern analysis that might have detected the geographic and financial anomalies. Risk aversion: raising concerns about a program designed to feed children — a sympathetic objective — carried political risk that approving questionable claims did not. Institutional loyalty: the officials who raised concerns found them minimized or dismissed by supervisors invested in the program's success. Procedural formalism: applications that met the formal documentation requirements were approved, because meeting formal documentation requirements was the definition of approval. Cultural normalization: the culture of the oversight office had been shaped by years of approving applications with minimal scrutiny. Capture of the expert class: the legal and accounting professionals who were part of the oversight system interpreted their roles as facilitating program participation rather than ensuring program integrity.
The result was $250 million stolen from a program designed to feed children, by people who understood how to navigate a system that rewarded procedural compliance over substantive review. No single failure was catastrophic in isolation. The aggregate of the seven layers was.
4.4 Parallel Patterns Across Jurisdictions
The structural features of the Feeding Our Future case are not unique to Minnesota or to the Child Nutrition Programs. They are the predictable output of the incentive structure described throughout this essay, applied to federal-state partnership programs that distribute large amounts of money through bureaucratic pipelines with limited end-to-end accountability.
In California, the Medi-Cal program — the state's Medicaid program — has consistently generated some of the highest improper payment rates of any state Medicaid program in the country. The California Department of Health Care Services has implemented numerous corrective action plans. The improper payment rate has fluctuated but has not been systematically reduced to acceptable levels. The structural features that produce the problem — provider-based billing with limited service verification, distributed administration through managed care plans with their own accountability gaps, and oversight capacity that is systematically inadequate relative to the scale of the program — have persisted through multiple administrations of both parties.
In Massachusetts, the Supplemental Nutrition Assistance Program has faced recurring challenges with Electronic Benefits Transfer card fraud — a category of fraud in which benefits are accessed at point-of-sale terminals in ways that suggest trafficking rather than legitimate food purchasing. The pattern has been documented in federal audit reports. The state response has involved corrective action plans, enhanced monitoring commitments, and periodic enforcement actions. The problem has not been resolved.
These are not red-state or blue-state problems. They are not unique to any particular political culture or administrative tradition. They are the common output of a common structural arrangement: large programs, distributed administration, weak end-to-end accountability, and incentive structures that reward process compliance over outcome verification. The political coloring of the state matters less than the structural architecture of the program.
PART V: WHY SMART PEOPLE ENABLE IT
The phenomenon documented in the preceding sections presents an intellectual puzzle that deserves more serious attention than it usually receives. The people who staff and lead the institutions described are not, by and large, stupid. Many of them are highly intelligent, well-educated, and genuinely motivated by a sense of public service. They passed competitive examinations, earned advanced degrees, and chose careers in public service over more lucrative private alternatives. How do people like this come to preside over systems that consistently fail to deliver what they promise?
The answer is not primarily psychological in the sense of diagnosing character defects. It is primarily structural in the sense of describing how the environment in which these people work shapes the choices that are rational for them to make.
Cognitive Dissonance and Motivated Reasoning
The most fundamental mechanism is cognitive dissonance. People who have invested significant effort in careers they experience as meaningful are strongly motivated to believe that those careers are, in fact, producing meaningful results. The alternative — acknowledging that one has spent twenty years doing work that is largely performative rather than productive — is psychologically costly in a way that most people will expend considerable cognitive effort to avoid.
Motivated reasoning is the cognitive tool by which this cost is managed. The official who processes a complaint without adequate investigation does not think of himself as failing the person who filed the complaint. He thinks of himself as correctly applying the available resources to the prioritized matters on his docket. The prosecutor who declines the complex case does not think of herself as protecting her conviction rate. She thinks of herself as exercising sound judgment about where prosecutorial resources can be most effectively deployed. The supervisor who issues a generous performance rating to a mediocre employee does not think of herself as enabling underperformance. She thinks of herself as being a supportive manager who understands that ratings affect people's lives and careers.
Each of these rationalizations is plausible on its face. Each is also, on careful analysis, a sophisticated piece of motivated reasoning that protects the individual from confronting the real consequences of her choices. The complaint that wasn't adequately investigated involved a real person with a real problem that did not get resolved. The case that wasn't prosecuted involved real harm that went unaddressed. The mediocre employee whose generous rating extended his career continued to occupy a position where he could have been replaced by someone who would have served the public better.
Career Incentives That Reward Equilibrium Over Disruption
In virtually every large public institution, the career outcomes of officials who maintain equilibrium are superior, on average, to the career outcomes of officials who disrupt it. This is not a universal rule — there are institutions where genuine accountability is valued and rewarded, and there are individuals who have succeeded by insisting on it. But the aggregate pattern is clear enough to constitute a structural reality.
The official who identifies and escalates a problem creates work for her supervisors, raises questions about the institution's performance, and potentially generates adverse external attention. Even if the problem is real and the escalation is appropriate, the official who raised it is now associated with bad news. The official who encounters the same problem, concludes it is within normal parameters, and closes the matter without escalation creates no such disruption. Her supervisors are not bothered. No external attention is generated. The matter is resolved in a way that requires nothing of anyone.
Over the course of a career, the signals this sends accumulate. Officials who look for problems find them and are held responsible for the disruption their findings cause. Officials who do not look for problems do not find them and proceed smoothly. The rational career strategy, in most public institutions, is not to look too hard.
The Comfort of Procedural Virtue
Perhaps the most psychologically interesting feature of institutional life in the administrative state is the genuine sense of virtue that procedural compliance generates. The official who has followed every required procedure, filed every required report, consulted every required stakeholder, and documented every required step has, by the standards her institution recognizes, done her job correctly. She can go home feeling that she has done her duty.
The fact that the procedure produced no useful outcome is, in this framework, not her responsibility. She did not design the procedure. She is not the one who should evaluate whether the procedure is fit for purpose. Her responsibility was to follow the procedure, and she followed it. The gap between what the procedure accomplished and what was needed is a policy problem, not a performance problem.
This ethical framework — in which the official's duty runs to the institution's rules rather than to the outcomes those rules were designed to produce — is not obviously wrong in philosophical terms. Institutions cannot function if officials feel free to deviate from authorized procedures whenever they judge the procedures inadequate. The rule of law requires a degree of rule-following that is independent of outcome evaluation.
But in practice, the procedural virtue framework resolves the cognitive dissonance of institutional failure in a way that allows that failure to perpetuate indefinitely. The official who believes she has done her job correctly when she has followed the procedure has no particular motivation to improve the procedure. The institution that evaluates performance primarily through procedure compliance has no reliable mechanism for detecting outcome failure. The system is stable precisely because the people in it have found a way to feel good about outcomes they are not actually producing.
Groupthink and Institutional Loyalty as Psychological Stabilizers
The final mechanism is the social function of institutional loyalty in maintaining individual psychological stability. Working in an institution that is failing its public mission is potentially deeply uncomfortable for people who entered public service with genuine intentions. The discomfort can be managed in several ways: exit, voice, and loyalty, in Albert Hirschman's formulation.
Exit is expensive. A career built in the public sector is not easily transferred to private employment, particularly for mid-career officials in specialized roles. Voice is risky, for the reasons described above. Loyalty — the investment of one's sense of identity and purpose in the institution rather than in external standards of performance — is comfortable, socially reinforced, and psychologically sustainable.
The official who identifies deeply with her institution can interpret evidence of institutional failure as external attack rather than internal critique. Critics of the institution are adversaries, not evaluators. The people complaining about institutional performance don't understand the constraints we operate under, don't appreciate how hard we work, don't have access to the full picture. This framework is available at every level of institutional hierarchy, from the front-line employee who resents public criticism of her agency to the senior official who responds to oversight findings with defensiveness rather than accountability.
Groupthink — the suppression of individual critical judgment in favor of group consensus — is the cognitive mechanism by which this loyalty manifests in institutional decision-making. Groups that are cohesive, that face external pressure, and that have strong norms against internal dissent tend to make decisions that confirm the group's existing beliefs rather than decisions that are best supported by available evidence. The decision to close a case without investigation, to decline a complaint, to issue a favorable performance rating, to recommend against escalation — in each case, the individual's independent judgment is influenced by the awareness of what the group expects, what the supervisor wants, and what the institutional culture rewards.
PART VI: UN-PADDING THE SHOVEL
Conclusion and Reform
6.1 The Return to Tangible, Site-Level Accountability
The central pathology of the modern administrative state is the substitution of process for outcome as the operative definition of performance. Reversing this substitution requires rebuilding accountability systems around what the institutions actually produce for the people they serve, rather than around what they document about their own activities.
This is harder than it sounds. Outcomes are often difficult to measure, especially in the short term, and the causal relationship between specific official actions and downstream outcomes is rarely clean. The police department that invested heavily in community relations may see crime increase due to factors entirely outside its control. The social service agency that followed correct procedures may see its clients deteriorate due to circumstances it could not have addressed. Simple outcome metrics — crime rates, benefit fraud rates, student test scores — often measure the characteristics of the population being served as much as they measure the quality of service.
None of this, however, is a reason to persist with purely process-based accountability. It is a reason to design accountability systems with care, to use multiple indicators rather than single metrics, and to build in the capacity for qualitative evaluation alongside quantitative measurement. Site-level accountability — the evaluation of specific officials in specific positions based on what their specific units actually accomplish, rather than on institutional averages — is the direction that serious accountability reform must take.
The model is not obscure. It exists in embryo in the best elements of performance management in high-functioning public agencies and in the accountability structures that have been demonstrated to work in specific institutional contexts. Emergency medical services systems that track outcomes by unit and by shift. Tax enforcement offices that evaluate auditor performance by audit quality, not audit volume. School districts that measure teacher effectiveness by student learning growth, not by credentials or seniority. These are imperfect examples of what becomes possible when the question asked is 'what did you actually accomplish?' rather than 'did you follow the procedures?'
6.2 Ending the Fresh Start Culture
The fresh-start norm — the systematic institutional refusal to maintain and act upon longitudinal knowledge about the actors and situations it encounters repeatedly — is correctable through deliberate design. The information needed to identify patterns already exists in most institutional contexts. What is lacking is the analytical infrastructure to aggregate it and the institutional will to act on what aggregation reveals.
Mandatory pattern recognition requirements — requiring agencies that process complaints, claims, or applications from the same actors repeatedly to flag and evaluate patterns before taking action on individual matters — would be a direct structural remedy. The technical infrastructure for this is not exotic. Pattern recognition across case records is a solved problem in information technology. The barrier is institutional rather than technical: the fresh-start norm is an institutional preference, not a technological constraint, and changing it requires changing the institutional incentives that make it preferable.
Cross-agency information sharing — allowing courts, law enforcement agencies, and administrative bodies to access each other's records of prior interactions with the same parties — would enable pattern detection at a scale that individual agencies cannot achieve alone. The legal barriers to such sharing are real and require careful navigation, particularly where privacy interests are significant. But the current regime, in which each institutional actor starts fresh with every encounter and no one is accountable for the aggregate pattern, is not privacy-protective in any meaningful sense. It is accountability-protective, and primarily protective of institutional actors rather than citizens.
6.3 Structural Reforms Worth Considering
Performance-linked compensation in the public sector is not a novel idea, and its history is mixed. Many performance pay experiments in government have failed, primarily because they grafted performance-pay structures onto institutions whose underlying culture and evaluation systems were not designed to support honest performance assessment. Supervisors who have never been trained to distinguish between genuinely excellent and merely adequate performance, and who face social costs for issuing low ratings, will issue uniformly high ratings regardless of whether a bonus is attached to them. Performance pay in this environment produces no accountability benefit and a significant cost in the form of across-the-board bonuses.
The prerequisite for meaningful performance-linked compensation is an honest evaluation system — one that forces genuine distribution of ratings, that is subject to external audit, and that creates real consequences for supervisors who issue ratings that are systematically inconsistent with independent performance measures. This is a cultural transformation, not a structural tweak, and it cannot be accomplished by policy change alone. But the cultural transformation cannot proceed without the structural changes that make honest evaluation the path of least resistance rather than the path of greatest resistance.
Qualified immunity reform has attracted bipartisan support in the past decade, though legislative action at the federal level has not materialized. The practical argument for reform — that the current doctrine effectively immunizes a substantial class of government misconduct from civil accountability, and that this immunity has no convincing basis in the statute that was meant to create that accountability — is sound. The doctrinal argument that qualified immunity was invented without textual support and expanded beyond its original scope is documented by legal scholars and acknowledged by members of the Supreme Court.
A targeted reform — maintaining immunity for officials who rely on specific statutory authorization or specific judicial precedent but removing it for officials who simply assert that no prior case addressed their precise conduct — would restore the accountability function that the Civil Rights Act of 1871 was designed to create without exposing officials to liability for genuinely novel constitutional questions. Whether such a reform would survive the political dynamics of a divided Congress is a different question from whether it is sound policy.
Transparency and citizen-audit mechanisms — accessible public databases of institutional performance metrics, complaint outcomes, improper payment data, and qualified immunity invocations by agency and jurisdiction — would enable the kind of external accountability that internal mechanisms have failed to produce. The argument against transparency in government performance data is typically that it will be misunderstood, cherry-picked, or used for political purposes. This is true, and it is not a serious objection. The argument for transparency is that accountability requires information, and the people whose money funds the institutions and whose lives depend on their performance are entitled to information that allows them to evaluate what they are getting.
6.4 Final Reflection
The American constitutional order was built on a particular insight about human nature: that people placed in positions of authority will tend, unless checked, to use that authority in ways that serve their own interests rather than the interests of those they nominally serve. This insight was not pessimistic in a nihilistic sense. It was realistic in the sense of taking human beings as they actually are rather than as we might wish them to be. The constitutional design that followed from it — the system of competing ambitions, the structural friction between branches, the rotation in office, the transparency of local government — was an engineering solution to a human problem.
Two and a half centuries later, the engineering has been partially undone. Not by malice, for the most part, and not by conspiracy. By the ordinary logic of institutional self-interest, applied consistently over a long period of time, by people who were, in most cases, doing what was rational for them to do within the systems they inhabited.
The Pendleton reformers who created the civil service were trying to solve a real problem. The judges who invented qualified immunity were trying to protect officials from frivolous suits. The proceduralists who built administrative law were trying to create fairness and predictability. Each reform addressed a genuine dysfunction and created new ones in its wake. The new dysfunctions were not random. They followed a consistent direction: toward the insulation of the official from the consequences of her performance, and toward the substitution of process for outcome as the measure of public service.
The padded shovel is the result. The public servant who digs with it can lean on the handle without losing the tool. She can generate reports about the digging without the hole getting deeper. She can, in the extreme cases, fill in what others have dug without any mechanism detecting the substitution.
The citizen who needed the hole dug waits. The hole is not dug. The report says the hole was dug, or that digging the hole was determined to be outside the scope of the applicable procedures, or that the matter was reviewed and found to not meet the threshold for action, or that the file was forwarded to the relevant department for follow-up. In each case, the report is real. In each case, the hole is not.
A republic cannot survive indefinitely when its servants view the citizen primarily as an input to institutional processes rather than as the reason those processes exist. The Founders knew this. They built a system intended to prevent it. The system has drifted. Restoring Madisonian friction — real accountability, visible consequences, competition among the ambitions of those who hold power over citizens — is not nostalgia for a simpler time. It is the precondition for a republic that functions as one.
The shovel needs unpadding. The hole needs digging. The distance between the two facts is the measure of the work that remains.
— END —
