Poster Paper: Better Measuring the Efficiency of Nonprofits

Saturday, November 4, 2017
Regency Ballroom (Hyatt Regency Chicago)

*Names in bold indicate Presenter

Jason Coupet, University of Illinois, Chicago and Jessica Haynie, North Carolina State University


The performance of nonprofit organizations can be difficult to measure. When it comes to measuring efficiency, the nonprofit management literature often leans on administrative or “overhead” spending spending, or the ratio of administrative spending to total spending.

The problem with this approach is that administrative spending does not directly measure efficiency. Traditional managerial efficiency involves the “ability to turn inputs into outputs” (Luksetich & Hughes, 1997). Classical definitions of managerial (or Koopman’s) efficiency involves maximizing the ratio of outputs to inputs. While administrative spending can be considered an input, its sole use as a measure of efficiency ignores a) other critical inputs and b) and outputs.

A nonprofit organization can expend a large proportion of resources on administrative expenses or overhead and still be efficient, and can sometimes improve efficiency by increasing administrative or overhead expenses. A nonprofit, for instance, that hires an additional staff member whose marginal rate of productivity exceeds her marginal costs has improved efficiency by increasing its overhead or administrative expenses. Thus, an overhead ratio might well measure “top-heaviness”, but we argue it is not a useful measure of efficiency.

Using financial and performance data from the nonprofit housing sector in the Southeastern United States, we illustrate this point. We compute an overhead ratio for each organization (N=1346), then measure efficiency using two of the most well-known methodologies from managerial economics. First, we measure efficiency with Data Envelopment Analysis (DEA), a mathematical linear program combining all relevant output/input ratios into a single efficiency score for each organization.

 We then calculate efficiency with Stochastic Frontier Analysis (SFA), an econometric methodology estimating efficiency from the error terms of translog production functions. We then rank organizations from most efficient to least efficient and compare their rankings with Spearman’s correlations.

We find statistically significant correlation between DEA and SFA efficiency rankings, but we find that neither efficiency measure is related to the organizational rankings by efficiency ratio. We suggest that nonprofit scholarship measuring efficiency move toward more established and theoretically precise measures of efficiency.