The overhead myth: when low overhead is a bad sign
Judging charities purely by overhead punishes the ones investing in competence. What the research says, and what to look at instead.
Published July 16, 2026 · CharityIndex editorial team
By the numbers — from the CharityIndex dataset
88%
Median program expense ratio
89%
Share spending 75%+ on programs
Across 81,336 e-filed Form 990s with a functional expense split, the median program expense ratio is 88% (CharityIndex analysis of IRS data).
89% of those filings direct at least 75% of spending to programs — the threshold watchdogs generally call efficient.
For decades donors were told to pick charities with the lowest overhead — the least spent on management and fundraising. The instinct is understandable and, pushed too far, actively harmful: it starves organizations of the systems, salaries and oversight that make programs work.
What chronic underinvestment looks like
- Underpaid staff churning out the door with their expertise.
- No financial controls — which is how fraud and waste actually happen.
- "Creative" expense allocation to hit an arbitrary ratio, making the numbers less honest, not more efficient.
A more honest read
- Treat extremes, not averages, as signals: 40% overhead deserves questions, but so does an implausible 2% at a complex organization.
- Weigh the trend over several years, not one filing.
- Look at results-adjacent signals the 990 does carry: audited financials, an independent board, stable reserves, growing programs.
This is why CharityIndex's letter grade caps rather than worships efficiency: it combines the expense split with fundraising cost, balance-sheet health and accountability signals, so a charity can't win an A by starving its own infrastructure.