Chapter 11 — Key Takeaways

The federal bureaucracy is the institution most Americans interact with — through the IRS, Social Security, the VA, USPS, the FAA, the FDA, OSHA, and dozens of others — even though it is rarely the institution they think about during election season. The chapter's central points:

Scale and structure

  • The federal civilian workforce is approximately 2.1 million employees, roughly stable for decades. The military adds about 1.4 million active-duty plus 800,000 reservists.
  • The "shadow workforce" of ~5 million federal contractors and ~3 million federal-grant employees brings Paul C. Light's "true size" estimate to roughly 9 million people executing federal policy.
  • The executive branch contains 15 Cabinet departments, ~60 independent agencies, ~50 government corporations, and ~40 boards and commissions. Headcount is unevenly distributed: the VA is the second-largest department; the Department of Education is the smallest.
  • The federal government produces roughly 80,000 pages of Federal Register content annually and 3,000–4,000 distinct rules per year, of which 50–100 qualify as "economically significant" ($100M+ annual impact). The Code of Federal Regulations runs to roughly 200,000 pages.

The civil service

  • Pendleton Act of 1883 established merit-based hiring (in response to Garfield's assassination by a disappointed office-seeker). Initially covered ~10% of jobs; expanded over decades to cover most federal positions.
  • Hatch Act of 1939 restricts political activities of federal employees (no running for partisan office, no campaign solicitation, no using position for political purposes). Amended 1993 to permit some off-duty activity.
  • Civil Service Reform Act of 1978 (Carter) created OPM, the MSPB, the FLRA, and the Senior Executive Service.
  • Career employees have substantial due-process protections against adverse personnel actions. Federal unions can bargain over working conditions but not pay or benefits, and federal employees cannot strike (Reagan fired 11,000 PATCO controllers in 1981).
  • Schedule F (EO 13957, 2020; revived in modified form in 2025) allows agencies to reclassify "policy-determining" positions out of the competitive service. The most contested civil-service question of the 2020s. Defenders argue it restores presidential accountability over policy implementers; critics argue the "policy-determining" category is elastic and risks reverting to spoils-system dynamics. The chapter does not adjudicate.

Independent agencies and the unitary-executive challenge

  • Independent regulatory commissions (FCC, FTC, FEC, NLRB, SEC, CPSC, NRC, CFTC) have multimember boards with for-cause removal protections. Designed for insulation from short-term political pressure.
  • Humphrey's Executor v. United States (1935) provides the constitutional foundation: Congress can insulate "quasi-legislative" and "quasi-judicial" multimember commissions from at-will presidential removal.
  • Seila Law LLC v. CFPB (2020) limited Humphrey's: for-cause protection does not extend to single-administrator agencies exercising significant executive power. Collins v. Yellen (2021) extended Seila's logic to FHFA.
  • The unitary-executive theory holds that Article II's Vesting Clause requires presidential removal authority over all officers exercising executive power. Whether multimember commissions retain Humphrey's protection through 2030 is the major unresolved doctrinal question.

Rulemaking and the Administrative Procedure Act

  • The APA (1946) establishes notice-and-comment rulemaking: statutory authority → Notice of Proposed Rulemaking (NPRM) → public comment period → OIRA review (for economically significant rules) → final rule → judicial review under "arbitrary and capricious" standard.
  • Motor Vehicle Manufacturers v. State Farm (1983) established the hard-look doctrine: courts review whether the agency considered relevant factors, addressed counter-evidence, and offered a reasoned explanation.
  • OIRA (within OMB, in the Executive Office of the President) centralizes White House regulatory review under EO 12866 (Clinton, 1993, building on Reagan's EO 12291). Cost-benefit analysis required for economically significant rules; Value of a Statistical Life estimates run roughly $9–13 million.

The doctrinal shift of the 2020s

  • Chevron (1984) directed courts to defer to a reasonable agency interpretation when a statute was ambiguous. Cornerstone of administrative law for forty years.
  • Loper Bright Enterprises v. Raimondo (2024) overruled Chevron. Courts now decide questions of statutory interpretation using ordinary tools; agency expertise still matters on factual findings; existing Chevron-era rules are not retroactively void but are subject to renewed challenge.
  • The major-questions doctrine — crystallized in West Virginia v. EPA (2022) and Biden v. Nebraska (2023) — requires clear congressional authorization when an agency claims power over questions of vast economic and political significance. Has been used against the Clean Power Plan, the OSHA vaccinate-or-test mandate, the CDC eviction moratorium, and the Biden student-loan forgiveness program.

The accountability problem

  • The administrative state attempts to combine technical expertise (which elected officials cannot supply) with democratic accountability (which expertise alone cannot legitimate). The recurring complaint, from both sides, is that it gets the boundary wrong.
  • The progressive critique: agencies have been captured by regulated industries — the FAA's Boeing relationship leading to 737 MAX failures, the SEC's Madoff blindness, EPA's slowness on PFAS, the revolving door. Progressive reformers seek more agency capacity and less industry coziness.
  • The conservative critique: agencies have stretched their statutory authority to reach favored policy outcomes — the Clean Power Plan, the eviction moratorium, the OSHA vaccine-or-test rule, student-loan forgiveness. Conservative reformers seek less agency overreach and more policy-making by Congress.
  • Both critiques have merit. They are not contradictory: the same agency can be too powerful in some domains and too weak in others.

Implementation

  • Pressman and Wildavsky's Implementation (1973): even a simple program requires many consecutive bureaucratic approvals, each with some probability of failure. Multiplied together, the probability of overall success approaches zero.
  • The ACA's bureaucratic implementation (2010–14), particularly the Healthcare.gov launch failure of October 2013 and the Zients-led tech surge recovery, is the canonical modern illustration that bureaucratic competence — not just policy design — determines whether laws become functioning programs.
  • The U.S. Digital Service (2014) and 18F were institutional responses; both have continued under every administration since.

What the chapter does not say

  • The chapter does not adjudicate whether the administrative state is good or bad, whether Loper Bright was correctly decided, whether Schedule F should expand or be repealed, or whether the major-questions doctrine should apply broadly or narrowly. These are contested constitutional and policy questions on which serious people genuinely disagree. The chapter steel-mans both sides and equips you to evaluate the disagreement for yourself.