Star ratings are supposed to help donors find well run charities. A study of Charity Navigator’s system finds that charities cluster just above the cutoffs for a higher rating, partly by relabelling their spending, and that donors reward the extra star with more money.
The paper is by Jennifer Mayo of the University of Missouri. It appears in the February 2025 issue of the Journal of Political Economy Microeconomics.
Bunching at the cutoff
Charity Navigator, founded in 2001, is the most used charity evaluator in the United States. It converts each charity’s financial health and accountability into a raw score, then into stars: 90 and above earns four stars, 80 to 90 earns three.
Mayo collected the full set of rated charities from 2002 to 2019 through Charity Navigator’s own data feed, so she could use the exact figures behind each rating. If charities ignored the stars, scores should flow smoothly across the thresholds. They do not. In one example from the data, 6,697 observations had scores between 89 and 90, but 7,644 sat between 90 and 91.
Using bunching methods borrowed from tax research, she estimates that the response is equivalent to charities cutting their administrative spending in half. The clustering was significant only around the four-star threshold, and it became sharper once charities got used to a new methodology introduced in 2016.
What a star is worth
Because charities sort themselves around the cutoffs, a simple comparison of three-star and four-star charities would overstate the value of a star. Mayo instead used the June 2016 switch to a new rating formula, which changed the star rating of around 27% of charities overnight for reasons unrelated to their own behaviour.
Moving from three to four stars raised contributions by 6% the following period, according to the published results. In the working version of the paper, larger charities saw a bigger gain of 9%. Donors did not appear to punish charities that lost a star.
Real savings or relabelling?
Not all of the response reflects leaner operations. Charities can raise their score by recording administrative or fundraising costs as program spending, which the formula rewards. Comparing rated and unrated charities, Mayo finds traces of this kind of relabelling, and concludes that part of the bunching comes from misreporting.
She also shows misreporting is cheap. Some charities near the four-star line would have collected more donations by cutting fundraising enough to cross it, implying average costs of about 1% of total expenses.
The study’s conclusion is about design: ratings with sharp cutoffs invite gaming, so evaluators need to account for charity incentives, donors’ reliance on shortcuts, and the need to monitor what charities report.
Study Details:
- Title: Navigating the Notches: Charity Responses to Ratings
- Authors: Jennifer Mayo
- Journal: Journal of Political Economy Microeconomics
- Publication Date: February 2025
- DOI: 10.1086/732241