Last weekend, I attended a conference / brainstorm hosted by the American University, in partnership with the Urban Institute. Titled ‘Nonprofit Panel Data Symposium,’ the two day event attracted some of the top brains in the country, working in the nonprofit/ civil society sector. As a young scholar who is interested in the sector and whose work intersects with the nonprofit sector quite strongly, I had a big incentive to be there. While it is hard for me to share all the details that were discussed, I will offer a few insights, in the hope of answering the question I have just raised : Do we need accurate data about
nonprofits in the U.S.? 
The short answer to this question is : Yes. The longer answer is to explicate why this is so. I will use my own research as a starting point to answer this question. First off, there is no ‘accurate data’ out there. As one skeptical participant raised the question : ‘Is the data we have worth shit’? That is a good question to ask, to begin with. The fundamental reason is that any data that is publicly available usually comes with a few caveats and limitations – what is disclosed, how much of it is disclosed and how current is this data? While Form 990s tell us part of the story, they don’t tell us the whole story.
On a related note, while statistics can inform us about the trends in giving, the impact of certain policies on how nonprofits respond, they usually conceal as much as they reveal.
While there are some excellent initiatives that have been ongoing, in the field of nonprofit data collection – at Urban Institute, Lilly School of Philanthropy among other institutes, the felt need among researchers is that of an ongoing panel data-set. Such a dataset does not exist for the nonprofit sector. In my research, I came across this problem too. While the groups Iam interested in studying for my doctoral degree are American Muslims, there is a lot of ‘issue’ based data, but no real panel dataset – either in quantitative or qualitative terms – that can offer us an analysis of how the nonprofit sector has changed over time. This temporal element is important, in many contexts, to tell a story.
Research is usually about creating a narrative that is compelling and also ‘true.’ While the facts may vary over time, having a long time-duration analysis evens out the fluctuations that are inherent in any organizations’ lifecycle and they can be analyzed and understood more clearly, over a period of time. To this extent, yes, a panel data set is a good idea. As bad and imperfect as any data source is, having something tells us something. Not having any data and just speculating, based on pre-existing theories or worse still, on hypothesis is the worst form of research or scholarship. And I have seen this, when it comes to scholarship on Islam or even philanthropy. Better to let the data speak, rather than let one’s imagination run wild. As Danial Patrick Moynihan famously said “One may be entitled to one’s own opinions, but not one’s own facts.”
