Policy & Supporting Positions A snapshot of the current roster
Work in Progress - Beneath the cabinet

The US cabinet is about twenty people. The president's roster runs to thousands more.

The other pages count cabinet seats. But every US administration also hands out thousands of jobs below that: agency deputies, policy chiefs, confidential aides. The Plum Book is the goverment's own list of those positions. The question here is whether the pattern at the very top holds further down. Note: this is still work in progress and below is just initial findings.

2,907
Presidential & political posts filled now
30%
Of current positions sit vacant
~31%
Of those posts held by women (estimated)
How to read this page

The Plum Book records no race, ethnicity, religion or gender. The gender figures here are estimated from appointees' first names, an approximation that can't identify nonbinary people and is less accurate for names outside its reference data. Treat every percentage as rough, not exact. Full method at the bottom of the page.

Part I · The map beneath the cabinet

What the Plum Book is, and how big it is

The Plum Book used to be a printed volume, issued after each presidential election. That's changing. Under the PLUM Act of 2022 the print edition is being retired, and OPM now runs an online system that agencies update through the year. This analysis uses OPM's snapshot dated 15 June 2026. It lists every “policy and supporting position”, which is a mixture of presidential and political appointments together with some senior career and time-limited roles. It is not a straight split between the two.

Presidential & politicalPAS + PA + Schedule C · filled
2,907
Career senior serviceCareer appointments (CA, CG) · filled
2,454
Other excepted / time-limitedExcepted by statute, limited-term (XS, TA…)
666
Cabinet & cabinet-levelThe part the other pages count
~24
PAS Presidential appointment needing Senate confirmation. Cabinet secretaries and other senior officials.
PA Presidential appointment with no Senate vote.
Schedule C Policy and confidential aides just below the top, the largest of the three political categories.
XS / TA Excepted by statute and limited-term roles. Not necessarily the president's to choose, so kept seperate here.

These are currently filled positions, and they add up to 6,027 of the 6,846 filled. Another 819 filled positions carry no appointment-type code in the data, and 2,951 positions are vacant, about 30% of the 9,797 current ones. (The file also holds 5,980 records of departed incumbents, which are left out of every figure here.) By common estimates a president fills roughly 4,000 political jobs in total, including more than 1,200 that need Senate confirmation, which is broadly in line with the filled-plus-vacant totals.

Part II · The power gradient

The cabinet's gender split holds - and thins towards the top

Across roughly 2,900 currently filled presidential and political posts (Senate-confirmed, presidential and Schedule C), about a third are held by women, close to the ratio in the current cabinet. The balance is not even across ranks, though. The more senior the tier, the fewer women in it.

Estimated share who are women, by tierdashed line = 50%
Senior executivesExec Schedule + SES
20%
Senate-confirmedPAS - the top tier
24%
All political postsPAS + PA + Schedule C
31%
Fewer women the higher you climb, and no tier reaches half
Part III · Department by department

Where the balance is closest - and furthest

The overall figure hides a lot of variation. Education and Health & Human Services come closest to an even split among their political staff, while Justice, Commerce, Homeland Security and Defense are the most male. Shown are twelve departments with enough filled presidential/political posts to be worth comparing.

Estimated share of political appointees who are womendashed line = 50%
Education
49%
Health & Human Services
41%
Interior
33%
State
32%
Treasury
30%
Agriculture
29%
Housing & Urban Dev.
26%
Defense (sec. title: War)
26%
Labor
25%
Commerce
24%
Homeland Security
24%
Justice
20%
Part IV · How this was measured

The method, and what it can't show

These numbers come from a single data file, run through one imperfect method. Here's what was done, and where it falls down.

01

The source

OPM's PLUM Reporting data, the online successor to the printed Plum Book, dated 15 June 2026. It holds 15,777 incumbent records covering 9,797 current positions, since the file also logs earlier officeholders. Every figure here counts only current incumbents, meaning those with no recorded departure date. Departed incumbents are left out.

02

What counts as “political”

The gender charts use only clearly presidential and political posts: Senate-confirmed (PAS), presidential without confirmation (PA), and Schedule C. The vaguer categories that aren't necessarily the president's to choose, such as excepted-by-statute and limited-term roles, are left out, as are career appointments. Gender itself is inferred from first names (about 94% resolved), so it's an estimate rather than a count.

03

The estimate is not neutral

Name-based inference cannot see nonbinary people, and it resolves Anglo names more reliably than others. Because that unreliability tracks ethnicity, the residual error is not random, which is a reason to read these figures as rough rather than precise.

04

What's missing entirely

The Plum data hold no race, ethnicity, or religion, so those dimensions can't be told from this file at all. And it is one snapshot: it describes the current roster, not how any of this has changed over time.