Two cheers for Giving USA

On Tuesday I had the pleasure of attending the launch event for the Lilly School of Philanthropy’s annual Giving USA report. Each year, the Giving USA report has something of the feel of the announcement of Time’s annual person of the year, or the release of some big report on the economy by the Federal Reserve. It’s supposed to serve as a sort of barometer of the charitable sector, and its findings are greeted even by mainstream media with great interest.

I’m glad that the Lilly School leads the charge on compiling this data. It certainly seems like they are fulfilling a need for serious study of the charitable sector that is underserved by other academics and dwarfed, of course, by the amount of time and resources put into studies of other sectors of the economy.

Of course, the headline grabber this year—even if the folks at the Lilly School wouldn’t have wanted it to be this—was that charitable giving was down.

But, in truth, that report shows that giving was up in absolute dollars and dropped a mere 1.7% once you take inflation to account. Even at the event, Giving USA’s Managing Editor Anna Pruitt emphasized that when you look at the trend over two years (2017 saw a strong increase over 2016) instead of just one, the trend line remains quite positive. Between 2016 and 2018, total charitable giving increased 3.1% after adjusting for inflation.

Not very dramatic, is it? And therein lies the problem with these annual reports: translation. Clearly the media ran with a less than complete narrative, as the media does with any study.

But it’s worth pointing out that they were heavily primed for the narrative of downward trends by the past couple of years of griping from charitable sector interest groups about how the new tax law would cause charitable giving to dramatically decrease. The Lilly School itself jumped into this fervor, with a previous study claiming that the basic set of policy proposals that became the new tax law would decrease charitable giving by over $13 billion.

What all the data in the new Giving USA report seems to show is that this didn’t happen. Or if it did, it must have been counteracted by some unknown positive force that more or less balanced out the tax law’s supposed negative effects. Or maybe it just hasn’t happened yet, and we’ll see a much bigger decline next year.

We’ll never really know, because data of this sort emphatically does not speak for itself. This was clear in the presentation the Lilly School Dean Amir Pasic gave summarizing the report’s findings. Giving is a bit down, he said, and then the next slide mentioned three or four items—the tax law, stock market volatility, declining religiosity, and a few others—that might have driven the change.

But if we’re being honest, those are all basically guesses.

That’s another problem here: by its very nature as an annual report, the Giving USA document has to declare that each year there are dramatic “trends” and “changes” in giving that everyone in the nonprofit sector should pay attention to, even if that isn’t the case at all. It therefore becomes one small and possibly inadvertent driver in a sort of mass faith of perpetual change that’s common to many nonprofit conferences, reports, and seminars. “Everything’s new! Everything’s changing! Digital! Millennials! Data-driven management!”

I would find it amusing if at some point Giving USA just became up front about that. Perhaps their 2020 report could be dramatically reduced in size, and point out, for intellectual honesty’s sake, that whatever little bump or decline the data shows that year is almost certainly just random statistical noise—and that even if it wasn’t, your guess would be as good as mine about what’s driving it. It’s only possible to make meaningful evaluations of these trends after at least a three-to-five-year period, or after a dramatic (and therefore already obvious to most observers) macroeconomic event like the 2008 crash.

On balance it’s probably a good thing that in nonprofits, as in other sectors, there’s more data than ever before. And Giving USA is a part of that admirable growth. If you’re a nerd like many of us in the nonprofit sector, it’s just plain fun to read through the highlights and speculate about what the numbers are really saying. Plus, you can sound smart at board meetings if you memorize the juicier bits: “For the first time this year, Giving USA reported that individual giving dropped to only 68% of the charitable pie. Can I have $40,000 more to spend on foundation and corporate prospecting this year?”

But humans remain human, and as a rule we’re absolutely awful at interpreting data. We extrapolate from data trends that aren’t there. We react disproportionately to small shifts. We confuse human survey answers with human behaviors. We’re bad at determining causes of change, and even worse at accounting for the random variance that’s inherent in almost all data collection. We’re also good at discounting data that doesn’t conform to our preexisting narratives.

There’s nothing unique about nonprofits in this regard—the same problems are driving massive crises in almost every field that relies on quantitative methods.

So to fundraisers, foundation representatives, and nonprofit executives, I say this: it’s highly unlikely that the Giving USA report contains even a single “trend” that says something meaningful for your particular organization’s fundraising or grantmaking.

Instead of plowing through Giving USA’s 450+ page report, seek data that you can be confident is meaningful and actionable for your organization. That means focusing on benchmarks that are specific to your budget size, staffing level, or sector (and no, the huge “religion” sector, for example, from Giving USA probably isn’t specific enough). And it means focusing on your internal data and what it can say about how to improve your operations, from simple retention rate and lifetime giving calculations to predictive data modelling based on the behavior of your donors alone.

I urge you to be a data moderate. Think about the methodology of data presented to you. Don’t demand data that doesn’t exist or that is too costly to acquire. Create a culture of openness that permits human questioning or rethinking of what the data says. And above all don’t discount the human element in data collection and interpretation—it’s actually the most important part.

---