Bad data, or none?

At a phone-in press conference last Thursday, Clara Miller of the Nonprofit Finance Fund and Robert Otterhoff of Guidestar cautioned donors against using a charity’s Form 990 to determine whether it’s worthy of support.  By way of commenting on proposed revisions to the 990, Miller suggested that the IRS and everyone else:

  • Eliminate the emphasis on overhead expenses: focus on a charity’s outcomes, not its inputs.  Because administrative and fundraising expenses are assigned differently by different charities (not to say completely randomly), they’re neither revealing nor a good basis for comparison.
  • Eliminate the requirement to report salaries about $100,000, which suggests that paying a charity executive that much is per-se profligate
  • Acknowledge that larger organizations tend to have lower administrative costs, but that growing organizations should have high expenses because they’€™re investing.

Miller urged donors to stop treating the 990 as gospel, noting that it "penalizes nonprofits for having reserves, paying people adequately, and investing in systems."  And Ottenhoff, whose agency posts 990s online, nonetheless suggested that donors begin their investigations not with the 990 but with an  agency’s mission and programs.

No dispute here that the 990s offer a distorted and distorting picture of nonprofit operations.  But nature abhors a vacuum, and donors are a force of nature.  They’re not
going to stop using the "wrong" information until we’re willing to give
them a replacement.  And neither Miller nor Ottenhoff had one to suggest. 

It won’t do just to ridicule donors’ "obsession" with overhead
expenses, and to compare it unfavorably to the forms of evaluation used
in the for-profit world: "A businessperson would ask, ‘What do you
care what the overhead is if we’re delivering to our customers?’"  But donors mostly have no way of knowing whether, or what, nonprofits are
delivering to their customers.  We ask about cost because if the whole thing is a waste, we want it at least not to be much of one. 

How should donors secure information about outcomes to replace the 990 data about inputs?  Miller was vague: it would be useful to have data on "what did organizations actually do, capacity metrics or production metrics"–but those don’t belong in the 990.

Where can donors could go for information about a charity’s effectiveness?  Ottenhoff mentioned Guidestar, of course, and the Better Business Bureau, but ended up recommending a charity’s own Website–not because of what’s likely to be there, but because of what ought to be there:

An organization should be able to say how many people have we served, what was the benefit we provided–did they learn to read, did they get fed?–let’s quantify the benefits we’ve provided to the world.  Organizations should be able to connect the dots between mission and money with good numbers.  Let’s look at the bigger picture of what they’re trying to accomplish.  Let’s say they have a $1 million budget and employ 12 people, what is the output of that investment and what are you as a donor enabling them to do?

That’s the question, all right.  But don’t tell people to ask it if you can’t help them answer it.


The Nonprofiteer will take the long weekend to celebrate Thanksgiving.  Publication resumes Monday, November 26.  Happy holidays!


Tags: , ,

6 Responses to “Bad data, or none?”

  1. Holden Says:

    Howz about good ol’ ?

  2. John C McGee Says:

    What is the output of any investment in a nonprofit and what should a donor expect that investment to accomplish? Both valid questions but to get to that point you need both the tools and the training to provide the information.

    What I see is a miniscule attempt to provide training to the small and mid-size nonprofits. This effort is both in the area of the reporting tool and the process by which that data can be collected. Without an understanding of the objective and how to process the data only the large organizations will be in a consistent position to provide the date to answer the opening questions. Creating a site to post data that provides outcome measures will fall significantly short of its objectives if a sizable portion of the sector is not equipped to provide the requested information

    For example the I.R.S. has made much of their pending trainings for small and mid-size nonprofit organizations. With a total of 17 sessions currently scheduled with 200 seats per session, (3,400 total) the Service is offering training to a maximum of .319% of the entire registered pool of 501(c)(3) candidates if one uses as it total the figures provided by the Chronicle of Philanthropy in its April 5, 2007 edition.

    If one assumes that only 50% of those organizations are “small and mid-size” then the Service is making available seats for approximately .68% of the target audience. Nothing in these figures would suggest that the effort by the I.R.S. while admirable is aggressive outreach to that portion of the exempt organization picture that have few resources and less time to devote to the information needed than their larger counterparts.

    How does one expect the small and mid-size organizations to meet the evolving standards for reporting activity if those who have the means to help them do not take steps to help them?

  3. Nonprofiteer Says:

    Holden: researches, by your own account, only charities working in New York and Africa. The rest of the nation and the world need service, too. Also, though it’s clear to me that you’re approaching the subject with rigor, I don’t quite understand your methodology. We’re still in the infancy of the science of determining what constitutes charitable effectiveness. F’rinstance: your site links to a critique of bednets as an effective intervention. The critique argues that bednets aren’t such a great deal because relatively few children die of malaria anyway. But the real problem with malaria is that it destroys the functionality and productivity of the people who live through it; children who don’t die of malaria are nonetheless markedly less able to learn, work and prosper for the rest of their lives. Now we can have a debate about whether that’s as important as making sure children don’t just die flat-out, but it’s just one example of thousands of a point on which we can’t agree because we haven’t first agreed on what’s at stake. I salute every effort to quantify nonprofit results, and have no patience for groups that refuse to; but let’s make sure we’re actually measuring what’s being accomplished rather than some dramatic subset thereof.

  4. Nonprofiteer Says:

    Mr. McGee: Thanks both for highlighting the existence of the IRS program (of which I was unaware) and its pitiful size compared to the need it’s supposed to be fulfilling. This may be a place where the biggest philanthropies–which are, after all, imposing significant reporting requirements of their own–could make a useful investment: provide free training to every grantee nonprofit (plus some small fraction of non-grantee nonprofits) in how to assess effectiveness in certain kinds of social services and how to distill those assessments into a format readily comprehensible to prospective donors.

  5. Stephen Peelor Says:

    I agree that the 990 and other standardized forms of reporting do little to convey the outputs and impacts that nonprofits have. The first step will be to have such formats include output measures – how many people served, kids taught, patrons attended, etc. – then the outcomes/impact can come next.

  6. Nonprofiteer Says:

    That distinction you make between outputs and outcomes is very helpful–if prospective donors at least know how many people are being served and with what, they have a basis on which to choose to offer or withhold support, even before we’re able to tell them how effective those services are.

Comments are closed.

%d bloggers like this: