1. The Role of Ratings Agencies in the Market for Charitable Contributions: An Empirical Test Teresa P. Gordon, University of Idaho Cathryn L. Knock, University of Idaho Daniel G. Neely, University of Houston Present at the GNP Mid-Year Meeting 2008 Presentation at the 2008 AAA Government and Nonprofit Section’s Mid-Year Meeting In San Antonio, Texas

2. Objective of this Study To determine if the ratings promulgated by Charity Navigator are incrementally informative to donors in their giving decision. It is an empirical question whether the aggregation of public information by an independent information intermediary into an overall measure of organizational quality is impounded into the donor’s giving decision.

3. Prior Literature Donors base their giving decision on a variety of factors (Gordon & Khumawala 1999). The efficiency of the organization appears to be one of the factors considered by donors (Glazer 1994, Tinkelman 1999).

4. Prior Literature (Continued) Donors appear to be willing to view financial information to help them make their donation decision (Buchheit and Parsons 2006) The inclusion of positive accounting information in fundraising appeals influences donations (Parsons 2006)

5. Charity Navigator Established in 2001 Rates over 5,000 organizations ◦ Information drawn almost exclusively from Form 990 To be rated an organization must have more than $500,000 in public support and four years of Form 990 data.

6. Economic Models of Giving Steinberg (1986) Approach ◦ First difference specification Similar to Tinkelman (1999) ◦ Include controls for current and prior year change in fundraising expenses ◦ Change in prior year “price” Departures from Tinkelman (1999) ◦ Do not control for lead and lag values of the change in administrative expenses and program expenses ◦ Consider percent change not raw change

7. Economic Models of Giving (continued) The more popular levels analysis (Weisbrod and Dominguez 1986, Okten and Weisbrod 2000, Tinkelman 1999) ◦ Control for the natural log of prior year ending assets, prior year “price”, prior year fundraising expense In both the levels and change specifications, consider the current rating change, and split the rating change into positive and negative rating’s changes. Also control for the prior year rating level (change) for the level (change) analysis.

8. Hypotheses H1: A positive / (negative) change in rating by Charity Navigator is associated with a positive / (negative) percentage change in contributions to a not-for-profit organization. H2: The level of current year contributions is a function of the current year change in Charity Navigator rating and the prior year rating of the organization.

9. Sample Selection Start with 5,233 organizations listed on the Charity Navigator website as of June 20, 2007. Randomly sample 10% of the organizations (525 charities). Restricted sample to organizations with three sequential years of financial data, no change in fiscal year end, and reported fundraising expenses > zero. 411 organizations in final sample.

10. Percentage Change in Contributions between First and Third Years Table 4, Panel A Rating achieved makes a difference

11. Percentage Change in Contributions between First and Third Years Table 4, Panel B Rating changes make a difference

12. The Change Analysis

13. The Levels Analysis

14. Limitations and Conclusion Rating changes appear to be positively correlated with current donations Results are robust to levels and change analysis. Smaller charities not included in analysis Charity Navigator not exclusive rating agency for these organizations Conditions correlated with the direction of the rating change could be driving the result.