Do operating reserves stabilize spending by nonprofit organizations?

Research output: Contribution to journalArticlepeer-review


The public budgeting literature has a long and rich tradition that examines the role of budget stabilization funds as fiscal stabilizers for state and local governments during periods of declining revenues and deteriorating economic conditions. Similarly, nonprofit organizations may accumulate operating reserves that allow them to smooth out annual imbalances between revenues and expenses, especially when facing a fiscal shock. Agency theory, on the other hand, indicates that managers might use these reserves to enrich themselves at the expense of the organization. This article is a step toward addressing a gap in our knowledge by analyzing the implications of reserves on nonprofit spending in general and also on particular functions (program versus overhead spending). Using a long panel of data from 1995 to 2011 and controlling for sample selection bias, the empirical results suggest that operating reserves held by nonprofit organizations do reduce expense gaps during downturns, but the effect is small. The results also suggest that nonprofit managers value current spending more than reserving funds for the future. Further, operating reserves are not associated with agency problems as predicted by theory. The empirical results suggest that the current rule of thumb—that nonprofits ought to hold up to 6 months of operating reserves—is inadequate if these pools of savings are intended to maintain all spending at trend during poor fiscal times. If, however, reserves are intended to only offset trend deviations partially while alternative strategies are sought, then the current rule of thumb may be sufficient.
Original languageEnglish (US)
Pages (from-to)295-311
JournalNonprofit Management & Leadership
Issue number3
Early online dateNov 14 2017
StatePublished - Mar 1 2018


Dive into the research topics of 'Do operating reserves stabilize spending by nonprofit organizations?'. Together they form a unique fingerprint.

Cite this