Nearly One in Three OPM-Covered Federal Employees Was in the Excepted Service

The term excepted service sounds like a narrow exception to the federal government's ordinary workforce. The numbers show something much larger.

In May 2026, the latest complete governmentwide workforce file from the Office of Personnel Management contained 649,975 excepted-service employees. They represented 32.1% of the 2.03 million employees covered by that file. The competitive service accounted for 1,369,110 employees, or 67.6%. Another 6,637 employees were in the Senior Executive Service.

Nearly one in three covered employees was therefore in the excepted service before the administration transferred roughly 8,000 policy-influencing positions into the new Schedule Policy/Career category in June.

That sequence matters. The 2026 reform did not create a 650,000-person alternative civil service. It moved a much smaller group into a large, long-established category that already included nurses, physicians, attorneys, aviation-safety personnel, air traffic controllers, national-security employees and many other career workers hired under specialized systems.

The category is also not a synonym for political appointment. Only 2,018 employees in the May snapshot carried the Schedule C appointment code, 0.31% of the excepted-service total. By comparison, the Department of Veterans Affairs alone had 301,726 excepted-service employees.

The policy stakes are still real. Competitive and excepted positions can differ in how applicants are examined, how veterans' preference applies, what probationary or trial period governs an appointment and what appeal rights an employee has. The new Schedule Policy/Career rule changes removal and appeal procedures for covered policy-influencing positions. But none of those distinctions supports treating every excepted employee as political, unprotected or outside merit principles.

FederalHiringData reconstructed 25 years of OPM workforce records, analyzed recent personnel actions and compared the employee stock with current USAJOBS announcements. The result is a picture of two major career-workforce systems whose boundaries reflect mission, occupation and statute more than a simple divide between ordinary and exceptional jobs.

One federal workforce, three service categories

OPM's FedScope data definitions identify three position-occupied categories: competitive service, excepted service and the Senior Executive Service. That is the classification used here.

The competitive service is the federal government's default civil-service system. OPM describes competitive hiring as a process in which agencies rate applicants against qualification standards, apply veterans' preference where required and select from the resulting candidates. Agencies may also use authorized alternatives such as direct-hire authority without turning the position into an excepted-service job.

The excepted service covers positions that are specifically excepted from competitive-service examination or appointment procedures by statute, executive order or OPM action. It includes multiple schedules and many positions recorded simply as other permanent or nonpermanent excepted appointments. The reason for an exception can be occupational, operational, statutory or security-related.

The SES is separate from both. It includes executive positions above the GS-15 level and below presidential-appointment levels, with career, noncareer and limited appointment types of its own. Folding the SES into either competitive or excepted totals would obscure the source data.

These totals describe the workforce OPM publishes, not every federal worker. The file excludes the Central Intelligence Agency, Defense Intelligence Agency, National Geospatial-Intelligence Agency, National Security Agency, Office of the Director of National Intelligence, U.S. Postal Service and other listed entities or populations. State Department Foreign Service personnel have been excluded from FedScope coverage since March 2006.

Those omissions are especially important here. Several excluded intelligence agencies rely heavily on noncompetitive or excepted personnel systems, while the Foreign Service has its own statutory appointment structure. Assigning those absent employees to either side would create precision the public data does not contain.

The excepted share grew long before the 2026 reform

FederalHiringData classified OPM's annual September employment cubes from 2000 through 2024 using the published Type of Appointment codes. September 2025 and May 2026 use the newer Federal Workforce Data position-occupied field. The chart marks that source transition because a 25-year series should not hide a change in the underlying file.

In September 2000, 76.9% of the covered workforce was competitive and 22.7% was excepted. The excepted share reached 27.9% by 2005, 29.9% by 2010 and 31.3% by 2020. It was 31.9% in September 2024 and 32.1% in May 2026.

The long movement was not a straight line. Competitive employment rose during parts of the 2000s, and service shares shifted with hiring, separations, organizational changes, statutory authorities and changes in the populations OPM covered. The State Department Foreign Service coverage break is one example of why the line should not be read as a single policy's effect.

Still, the scale is unambiguous. The excepted service had more than 400,000 covered employees in 2000 and more than 600,000 by 2009. It was already close to one-third of the published workforce years before Schedule Policy/Career existed.

The new schedule belongs in that history, but it cannot explain it. The White House's official figure of about 8,000 transferred positions equals roughly 1.2% of the May 2026 excepted-service stock. Even the approximately 50,000 positions contemplated in the final rule's regulatory discussion would have been a minority of the existing category. The actual transfer announced in June was much smaller than that planning estimate.

Veterans Affairs accounts for nearly half the excepted workforce

Excepted employment is not evenly distributed. It is concentrated in agencies whose missions depend on clinical, transportation, legal, regulatory and security work.

VA employed 301,726 excepted workers in May, or 46.4% of the entire covered excepted-service population. The Department of Homeland Security followed with 86,641. Justice had 48,155, Transportation 45,064 and the Army 37,833.

Related research

Put this finding in context

All articles

Data investigations/August 24, 2026

The VA Is Exploring AI for Disability Claims. It Has 2,950 Fewer Claims Examiners.

VA's oldest claims backlog has fallen even as its covered claims-examining workforce declined 14% from September 2024 through June 2026.

Data investigations/August 21, 2026

The Veterans Health Administration's Federal Workforce Fell by 28,342. Its Care Forecast Kept Rising.

VA's health workforce contracted after late 2024 even as patient, visit and community-care demands remained large. The losses reached clinical and coordinating roles.

Large counts and high shares answer different questions. VA had both: 67.7% of its covered employees were excepted. Transportation's share was even higher at 85.5%, driven by occupations such as air traffic control and aviation safety. The Nuclear Regulatory Commission had the highest share among agencies with at least 1,000 covered employees, 96.8%, but a much smaller headcount.

VA illustrates why a binary label can conceal more than it reveals. Its workforce includes Title 38 and hybrid Title 38 health-care personnel, General Schedule employees, wage-system employees and other pay and appointment authorities. The public employment file does not cleanly identify every statutory Title 38 position.

FederalHiringData therefore used a strict, visible subset rather than claiming a comprehensive Title 38 count. Three clearly clinical VA pay plans - VN for nursing, VM for medical and dental, and VP for podiatrists and optometrists - contained 112,082 employees, all recorded as excepted. That subset included 89,316 nursing employees and 22,369 medical and dental employees.

It is not the whole story. VA also had 168,601 excepted employees on the General Schedule and 17,823 excepted employees in an administratively determined pay plan whose statutory basis cannot be isolated from the public field. The defensible conclusion is that clinical systems explain a substantial part of VA's excepted workforce, not that every VA excepted employee can be assigned to a single Title 38 authority.

Excepted service is occupational infrastructure

The occupation data makes the category's operational character clearer.

Among series with at least 500 employees, Aviation Safety was 97.8% excepted. General Attorney was 97.5%. Nursing Assistant was 95.0%, Air Traffic Control 92.6%, Social Work 91.8%, Medical Support Assistance 90.6%, Medical Officer 88.6% and Nurse 87.6%.

Attorneys provide a useful example outside health care. OPM's federal attorney hiring guidance says attorney positions are excepted under Schedule A because OPM does not examine for professional legal positions. Agencies must still follow merit-system principles, apply qualification standards and observe veterans' preference as far as administratively feasible.

That is an exception from the ordinary examining framework, not an exemption from all civil-service values.

The same caution applies across the ranking. A service label does not reveal a position's exact hiring authority, preference rules, trial period or appeal rights. Two excepted employees may have very different statutory protections. An occupation may also contain competitive, excepted and SES positions at the same time.

The largest excepted occupation counts reinforce the point. Nurses accounted for 92,978 excepted employees. Compliance Inspection and Support accounted for 48,411, Medical Support Assistance 36,149, General Attorney 35,729 and Medical Officer 30,953. Those are not marginal staffing systems. They are major parts of federal health care, enforcement, adjudication, safety and public service.

Schedule C is a sliver, not the definition

The appointment detail directly tests the common assumption that excepted service means political service.

In May 2026, 510,831 excepted employees - 78.6% of the category - were coded as other permanent excepted-service appointments. Permanent Schedule A accounted for 56,974 employees and nonpermanent Schedule A another 20,589. Other nonpermanent excepted appointments accounted for 50,671.

Schedule C, the category OPM describes as positions of a confidential or policy-determining character associated with noncareer appointments, contained 2,018 employees. That was 0.31% of the excepted total.

Schedule C should not be used as a proxy for every political appointee, and the table is not a complete political-appointment census. It is instead a decisive check on the reverse claim: the enormous majority of employees recorded in the excepted service were not Schedule C.

Both systems contracted during the 2025 personnel-action surge

The 2025 workforce disruption crossed the competitive-excepted boundary.

FederalHiringData grouped OPM accessions and separations by the employee's position-occupied category for January through May of 2024, 2025 and 2026. These are processed personnel actions, not necessarily unique people. They include more than one kind of accession or separation and should not be subtracted from headcount as if they formed a perfect stock ledger.

In the first five months of 2024, accessions exceeded separations in both systems. In the same period of 2025, competitive accessions fell 57.8% from the prior year while competitive separations rose 28.8%. Excepted accessions fell 46.9%, and excepted separations rose 49.3%.

By January-May 2026, accessions had recovered from the 2025 lows in both categories. Competitive accessions were up 32.3% from 2025, and excepted accessions were up 21.8%. But separations still exceeded accessions: by 3,963 actions in the competitive service and 2,463 in the excepted service.

This is evidence of a broad workforce shock and partial recruiting recovery, not a clean causal estimate for any one executive action. Personnel actions can be revised, and transfers, conversions and changing coverage complicate a direct reconciliation with monthly headcount. The comparison does show that the excepted service was neither insulated from the 2025 contraction nor uniquely responsible for it.

USAJOBS is more excepted-heavy than the employee stock

Current public recruiting provides a second, narrower view.

On Aug. 29, 2026, FederalHiringData captured 13,364 active USAJOBS announcements with a recognized service type. Of those, 7,661, or 57.3%, were labeled competitive service. Another 5,597, or 41.9%, were labeled excepted service. The remaining 106 were SES announcements.

The excepted share of announcements was about 10 percentage points higher than its share of covered employees. That difference is meaningful as a recruiting snapshot, but it is not a hiring rate. One announcement can advertise multiple positions or locations, remain open as a standing register, reach an applicant cap, or produce no hire. Employees and announcements use different denominators.

The comparison also cannot repair OPM's coverage exclusions. A public intelligence or Foreign Service announcement may appear in USAJOBS even when the corresponding employee population is absent from the workforce file. Service labels can be missing or applied at the announcement level rather than the individual position level.

For job seekers, the practical lesson is simpler: service type is not a minor field. A substantial share of current public announcements is excepted, and candidates should read each announcement's hiring path, who-may-apply section, appointment type, trial period and conditions of employment rather than assuming competitive-service rules apply to every listing.

Merit, preference, probation and appeals do not collapse into one label

Competitive-service hiring generally uses OPM qualification standards, competitive examining or an authorized alternative, and statutory veterans' preference. Excepted-service procedures vary with the authority. Some authorities require agencies to apply veterans' preference; others apply it differently or as far as administratively feasible. Merit-system principles continue to govern federal personnel decisions more broadly.

Probation terminology also differs. OPM's guidance for supervisors describes a probationary period for competitive-service appointments and a trial period for many excepted-service appointments. The duration and consequences depend on the position and authority.

Appeal rights require even more care. OPM's employee rights and appeals guidance explains that many competitive-service employees acquire adverse-action appeal rights after one year of current continuous service. Preference-eligible excepted employees can be covered after one year, while certain other excepted employees become covered after two years. Statutory exceptions and appointment-specific rules remain.

Service type therefore changes the legal starting point, but it does not answer every employee-rights question. Saying that all excepted employees lack appeal rights is wrong. Saying that competitive and excepted employees have identical rights is also wrong.

What Schedule Policy/Career changed

OPM published the final Schedule Policy/Career rule on Feb. 6, 2026. The administration describes the schedule as covering career positions of a confidential, policy-determining, policy-making or policy-advocating character. The rule says those positions remain nonpartisan and merit-based.

On June 3, President Donald Trump issued Executive Order 14410, transferring the positions listed in its appendix. The accompanying White House fact sheet put the affected population at about 8,000 positions.

That is the strongest authoritative count available for the actual transfer. It should not be replaced with the roughly 50,000 positions discussed as a potential scale during rulemaking. Nor should it be inferred from the May OPM data, which predates the transfer, or the incomplete June workforce file, which does not provide a clean new Schedule Policy/Career appointment code for this purpose.

The transferred positions remain career jobs, not Schedule C political appointments. According to the final rule, positions moved from the competitive service continue to use competitive hiring procedures, while positions already in the excepted service retain the applicable excepted procedures. What changes most sharply is accountability after appointment: covered employees are excluded from the adverse-action and performance procedures in chapters 43 and 75 of Title 5 and the associated Merit Systems Protection Board appeal route.

That is a consequential legal change for the affected workers. It is also a change involving about 1.2% of the pre-transfer excepted-service population, not evidence that the other 98.8% suddenly became policy staff.

Methodology and limitations

This analysis uses OPM's September FedScope employment cubes for 2000-2024, the newer Federal Workforce Data files for September 2025 and May 2026, January-May personnel-action files for 2024-2026 and a bounded current USAJOBS capture on Aug. 29, 2026.

May is the latest complete month suitable for a cross-agency ranking. OPM reports that the June 2026 release omits about 83,000 Defense records across 32 components. Using June would make governmentwide, agency and occupation comparisons look more current while making them less accurate.

Historical records through 2024 are classified from exact Type of Appointment codes: competitive codes 10, 15 and 20; excepted codes 30, 32, 34, 35, 36, 38, 40, 42, 44, 45, 46 and 48; and SES codes 50, 55, 60 and 65. Modern records use OPM's position-occupied category. Unspecified or invalid records remain separate.

The coverage exclusions mean this is an OPM-covered workforce analysis, not a complete census of everyone employed by the federal government. The VA pay-plan analysis identifies only clinical categories visible with confidence. Personnel actions are actions, not unique employees. USAJOBS announcements are not vacancies or hires.

Within those limits, the central finding is durable. Excepted service is not a small political edge around the competitive civil service. It is a nearly 650,000-person set of career and noncareer systems embedded in the government's delivery of health care, transportation safety, law, enforcement and national missions.

The 2026 reform changed the status and appeal path of a consequential group of policy-influencing employees. It did not create the broader category, explain its long growth or turn its nurses, doctors, attorneys, controllers and inspectors into political appointees.

Understanding that distinction does not minimize the new policy. It makes the debate accurate enough to have.