Predictive Modeling of the Non-Profit Sector in the US

The Non-Profit Sector contributes almost $1 trillion to the US economy, representing 5.4% of GDP, and generating over 12 million jobs in 2017. Yi (2010) suggests that a better understanding of the factors that affect fundraising should be of great interest to policy makers, and fundraisers. However, the workings of the sector are subject of much debate. Matsunaga, Yamauchi and Okuyama (2010) relate its size to the Theory of Government Failure. Sokolowski (2013) proposes that government funding does have a positive effect on revenues. Curry, Rodin and Carlson (2012) suggested they swing with GDP, but, Berman, Brooks and Murphy (2006) contend that macroeconomic variables do not affect short-run dynamics. List (2011) found that non-profit revenues react more to economic upswings than downturns. And the National Philanthropic Trust (2016) relates ups and downs to certain events and public awareness. Wallace (2016) points to the fact that predictive modeling has focused big-donor analytics, aimed at the identification of potential donors. We set out instead to define a working model. After locating complete time series for an emblematic segment, the environmental cause, Factor Analysis allowed us to pinpoint independent variables. We found that Non-Profit Revenues (NPR) depend largely on Public Awareness, as measured by TV coverage, and Disposable Personal Income (DPI), specifically: NPR = -4401.542 + 528.327(DPI) +23.121(TVCoverage) +