eTopical Precious Finds
A program efficiency ratio measures one thing: the share of a nonprofit’s total expenses that its accountants assigned to program activities rather than to management or fundraising. It measures an accounting allocation. It does not measure whether the programs worked.
That gap between what donors think the number says and what it actually says has been the subject of an organized correction campaign by the three largest charity evaluators for more than a decade.
Where the number comes from
Every 501(c)(3) filing a full Form 990 completes Part IX, the Statement of Functional Expenses. Part IX runs four columns: total expenses in column A, then program service, management and general, and fundraising in columns B, C and D. Line 25 totals each one.
The program efficiency ratio divides column B by column A. A charity reporting $8.5 million of $10 million in program expenses posts an 85 percent ratio.
Column assignment is a judgment call, and that is the whole problem. A staffer who spends mornings running a program and afternoons writing grant reports gets split across two columns by someone’s estimate. A mailing that describes a problem and asks for money can be allocated partly to program and partly to fundraising. Two honest organizations doing identical work can report meaningfully different ratios.
The evaluators told donors to stop using it
In June 2013, the heads of the three organizations donors most often consult published a joint open letter. Art Taylor of the BBB Wise Giving Alliance, Jacob Harold of GuideStar, and Ken Berger of Charity Navigator signed it together.
Their letter calls the percentage of expenses going to administrative and fundraising costs “a poor measure of a charity’s performance.” It asks donors to weigh transparency, governance, leadership, and results instead. It concedes that overhead has a narrow legitimate use, in rooting out fraud and poor financial management. And it argues that many charities should spend more on overhead, not less.
The letter closes: “The people and communities served by charities don’t need low overhead, they need high performance.”
The three followed up in October 2014 with a second letter, this one addressed to nonprofits, asking them to publish performance data and stop spotlighting financial ratios in their own marketing.
The starvation cycle
Ann Goggins Gregory and Don Howard of the Bridgespan Group named the underlying dynamic in “The Nonprofit Starvation Cycle,” published in the Stanford Social Innovation Review in 2009.
Their account runs in three steps. Funders hold unrealistic expectations about what running an organization costs. Nonprofits feel pressure to meet those expectations. So they underspend on infrastructure and underreport what they do spend, which confirms the funders’ original assumption and resets the cycle one notch tighter.
The evidence they assembled is specific. The Nonprofit Overhead Cost Study examined more than 220,000 Forms 990 and surveyed roughly 1,500 organizations with revenue above $100,000. More than a third reported no fundraising costs at all. One in eight reported no management and general expenses. Among four youth-serving nonprofits Bridgespan studied directly, reported overhead ran 13 to 22 percent while actual overhead ran 17 to 35 percent.
For comparison, Gregory and Howard note that overhead across for-profit industries averages about 25 percent of total expenses, and that no service industry reports average overhead below 20 percent. Meanwhile none of the government contracts held by those four nonprofits allowed more than 15 percent for indirect costs, and foundation allowances averaged 10 to 15 percent.
An organization reporting 8 percent overhead is not necessarily disciplined. It may be an organization that has learned what number to report.
What a low ratio can actually mean
Underreported allocation is one explanation. Three others recur.
Deferred infrastructure is common. An organization running on outdated case management software and unreplaced laptops posts excellent ratios until the year something breaks.
Suppressed pay is another. Salaries for finance and operations staff land in management and general. Holding those salaries down improves the ratio and raises turnover in the roles that keep an organization solvent.
Subsidy by another funder also distorts the picture. A nonprofit with donated office space or a parent institution absorbing its accounting reports a ratio that reflects someone else’s balance sheet.
What the standards actually require
The BBB Wise Giving Alliance still publishes numeric thresholds in its Standards for Charity Accountability, and knowing their exact wording prevents most misreadings.
Standard 8 asks a charity to “spend at least 65% of its total expenses on program activities.” Standard 9 asks it to “spend no more than 35% of related contributions on fund raising.” Related contributions is a defined term covering donations, bequests, special event income, federated campaigns, donated goods and services, and foundation and government grants. It is not 35 percent of total expenses, and the distinction changes the math substantially.
Both standards carry an extenuating-circumstances process. A newly created organization, one with heavy donor restrictions, or one working on a cause carrying public stigma can explain a miss rather than simply fail.
Two further standards get less attention and say more. Standard 10 asks charities to avoid accumulating unrestricted net assets exceeding three times annual expenses. Standard 13 flags joint-cost allocation, noting that allocating more than 50 percent of a fundraising appeal’s cost to program services will likely trigger a closer review. Standard 13 exists because the ratio is gameable, and the people who publish the ratio know it.
Charity Navigator has also moved well beyond its original financial-ratio system. Its current Encompass rating draws on roughly 45 to 50 metrics across four beacons, scored into a 0 to 100 percent result and translated into stars. The program expense ratio survives as one metric inside the Accountability and Finance beacon, and most of its financial metrics use three-year averages rather than a single filing.
Better questions than the ratio
What outcome does the organization claim, stated in terms that could turn out false? Who measures it? Does it publish results when they disappoint? Does its Form 990 show reserves sufficient to survive a bad year? Are the people affected by the work involved in shaping it?
Those questions take longer than reading one percentage, which is why the percentage persists. GiveWell’s experience is instructive on that point: the back page of the 2013 letter notes that its 2011 recommended charities carried higher overhead, 11.5 percent, than the organizations it reviewed and declined, at 10.8 percent.
Donors comparing organizations in a single field will find the work easier than comparing across fields, since cost structures differ enormously between direct service and research. Several nonprofits publish comparative material for their own sectors. Fight For A Living Wage, a nonpartisan grassroots 501(c)(3) whose stated mission is ensuring every American who works full-time can afford the basics, publishes one such list of groups working on poverty in the United States.
Read a Form 990 alongside any such list. The IRS makes them publicly available through its Tax Exempt Organization Search, and Part IX takes about ten minutes to work through.
