OpenFeds Analysis

Federal Workforce Trends: Is Government Right-Sized Now?

The federal workforce is the smallest it's been per-capita since before World War II. After eight decades of expansion punctuated by occasional drawdowns, are we finally at the right number — or have we cut too deep? The historical data provides essential context.

Sources: OPM, BLS, GAO, CBO, historical census data·Updated: July 2026

📜86 Years of Federal Workforce Size

Every generation thinks its government is uniquely bloated. The data tells a more nuanced story:

1940
1.04M(7.9 per 1,000 pop.)

Pre-WWII baseline

1945
3.82M(27.3 per 1,000 pop.)

WWII peak — wartime mobilization

1950
1.96M(12.9 per 1,000 pop.)

Post-war drawdown

1960
2.4M(13.3 per 1,000 pop.)

Cold War expansion

1970
3M(14.7 per 1,000 pop.)

Great Society peak; Vietnam era

1980
2.88M(12.6 per 1,000 pop.)

Post-Vietnam stabilization

1990
3.13M(12.5 per 1,000 pop.)

Cold War final years

1995
2.9M(11 per 1,000 pop.)

Clinton "Reinventing Government" cuts begin

2000
2.78M(9.9 per 1,000 pop.)

Post-reinvention stabilization

2010
2.78M(9 per 1,000 pop.)

Post-9/11 DHS creation; census year

2016
1.87M(5.8 per 1,000 pop.)

Obama-era (executive branch only)

2020
1.93M(5.8 per 1,000 pop.)

Pre-pandemic baseline

2024
2.07M(6.1 per 1,000 pop.)

Peak — pandemic/IRA hiring surge

2026
1.81M(5.3 per 1,000 pop.)

Post-DOGE restructuring

📈The Growth That Made Cuts Inevitable

Between 2016 and 2024, the federal workforce grew by nearly 200,000 employees — a 10% increase in eight years. This wasn't organic growth matching population increases. It was driven by legislation that created positions without asking whether existing capacity was sufficient.

The Inflation Reduction Act alone planned for 87,000 new IRS employees. The infrastructure bill added thousands of program managers. Pandemic-era emergency hiring became permanent. Each program had its own justification, but collectively they produced a federal workforce that was larger than it had been since the 1990s — without measurably better service delivery.

2016–2020
DHS expansion, VA hiring surge, census preparation
+55K
+14K/year
2020–2022
Pandemic response hiring (CDC, FEMA, SBA), vaccine distribution
+62K
+31K/year
2022–2024
IRA-funded IRS surge (87K planned), infrastructure bill staffing, climate programs
+83K
+42K/year
2024–2026
DOGE restructuring, hiring freeze, DRP, RIFs
-260,000
−130K/year

🌍How Does the U.S. Compare Globally?

Federal employees per 1,000 population — an imperfect but useful comparison:

United States (2026)
5.3
United States (2024)
6.1
United Kingdom
8.2
Canada
8.9
Germany
6.1
France
13.4
Australia
6.8
Japan
5.8

Note: These figures count federal/national government employees only, not state/local. The U.S. delegates far more to state and local government than most countries, so direct comparison understates total U.S. public employment. Still, the trend is clear: the U.S. runs a relatively lean federal government by international standards — and post-DOGE, it's among the leanest in the developed world.

📋The Right-Sizing Scorecard

Is the government “right-sized” now? The answer depends on what you measure:

Federal employees per 1,000 population

Pre-DOGE
6.1
Post-DOGE
5.3

Lowest since pre-WWII

Manager-to-worker ratio

Pre-DOGE
4.1:1
Post-DOGE
3.2:1

Improving but still above private sector

IT systems per federal employee

Pre-DOGE
1.2
Post-DOGE
1.4

Technology should enable smaller workforce

Average age of federal worker

Pre-DOGE
47.3
Post-DOGE
44.8

Younger workforce after senior DRP exits

SSA claims processing time (days)

Pre-DOGE
42
Post-DOGE
68

Concerning — some cuts hit essential services

VA healthcare wait time (days)

Pre-DOGE
21
Post-DOGE
24

Modest increase — VA was largely protected

IRS phone wait time (minutes)

Pre-DOGE
28
Post-DOGE
47

Noticeable degradation in tax season

🎯Our Assessment

The federal workforce needed restructuring. That's not partisan — it's math. A 10% headcount increase over eight years with no measurable improvement in service delivery is the definition of bloat. Twenty-three percent of eliminated positions were already vacant. The manager-to-worker ratio was worse than any Fortune 500 company.

But right-sizing is not the same as across-the-board cutting. The data shows a clear pattern: structural improvements in workforce efficiency alongside pockets of genuine service degradation.

The cuts that made sense:

  • Eliminating long-vacant positions (23% of cuts)
  • Flattening management layers (manager ratio from 4.1:1 to 3.2:1)
  • Consolidating redundant agency functions
  • Reversing hiring surges that were never meant to be permanent (IRA, pandemic response)

The cuts that didn't:

  • SSA field offices serving elderly Americans with no internet access
  • VA claims processors already working backlogs
  • GSA facility staff cut weeks before needing more facility capacity
  • Wildfire response teams before wildfire season

The bottom line: the federal government is closer to right-sized than it was in 2024. But “right-sized” is a process, not a destination. Some overcorrections will need fixing. The important thing is that we're finally having a data-driven conversation about what government should look like — instead of just assuming every position is essential because it exists.

📚The Clinton Precedent

This isn't the first time the federal workforce was dramatically reduced. Between 1993 and 1999, the Clinton administration's “Reinventing Government” initiative eliminated approximately 350,000 federal positions — proportionally similar to the DOGE reductions.

At the time, critics predicted catastrophe. Federal employee unions protested. Media coverage was overwhelmingly negative. Sound familiar?

What actually happened: government services continued. The deficit turned into a surplus. The economy boomed. And today, almost no one argues those cuts should be reversed. The positions eliminated in the 1990s are now universally regarded as having been unnecessary.

History doesn't repeat, but it rhymes. The question isn't whether 256,000 cuts will be remembered as transformative or destructive — it's whether the execution was thoughtful enough to achieve the former. The data so far suggests: mostly yes, with notable exceptions.