The Impact of Alternative Fundraising Schemes on Charitable Giving in Germany

Researchers:
Steffen Huck
Andrew Shephard
Fieldwork by:
Location:
Munich
Sample:
22,512 recent opera attendees
Timeline:
2006 - 2006
Outcome of interest:
  • Charitable giving
Intervention type:
  • Information
  • Targeting
Partners:

Nonprofits raise large amounts of money through fundraising and charitable giving every year. However, little is known about the best practices for attracting donations. Researchers partnered with the Munich Opera House to conduct a randomized evaluation measuring the impact of a large-scale mail fundraising scheme on the frequency and amount of donations. While neither the presence of a lead donor nor differences in matching schemes affected the response rate to the mailer, having a lead donor nearly doubled the average amount donated.

Policy issue

In 2017, Americans gave US$410 billion to charities, and 72 percent of charitable giving came in the form of individual donations.1 Despite the fact that these organizations often depend on the support of individual donors to finance their mission, there is minimal evidence on the relative effectiveness of alternative fundraising schemes. Understanding what motivates people to donate and how much they give is crucial to improving non-profits’ financial sustainability.

Context of the evaluation

In 2006, the Bavarian State Opera house was engaged in a social youth project entitled “Stück für Stück.” The project financed small workshops and events for schoolchildren with disabilities or from disadvantaged areas. These served as a playful introduction to the world of music and opera, and was a key activity for the Opera’s mission to encourage appreciation for the operatic art form in future generations.

Similar to fundraising drives by aid charities, the Opera chose to elicit donations for "Stück für Stück" through individual mail solicitations sent to their previous attendees. The average recipient of the Opera’s solicitation typically purchased six tickets in the year prior to the mail out, at an average cost of EUR 86 per ticket. Around 41 percent of recipients resided in Munich, where the opera house is located.

Woman reads papers at her desk

Details of the intervention

Researchers worked with the Bavarian State Opera to organize a mail out designed to shed light on the efficacy of alternative fundraising schemes in eliciting donations for "Stück für Stück."

A total of 25,000 recipients were randomly selected from the opera’s database of customers who had purchased at least one ticket in the twelve months prior to the mailing. Non-German residents, corporate donors, and formally titled donors were removed from the initial pool, resulting in a total of 22,512 addressees.

Each addressee was randomly assigned to receive one of six mailing programs:

  1. Comparison: A straightforward request for a donation.

  2. Lead Donor: Information about an anonymous lead donor who had committed to providing EUR 60,000.

  3. 50 Percent Matching: An explanation of a matching scheme in which an anonymous donor would donate 50 cents for each euro donated by the recipient (up to a total of EUR 60,000).

  4. 100 Percent Matching: An explanation of a matching scheme in which an anonymous donor would donate an additional euro for each euro donated by the recipient (up to a total of EUR 60,000).

  5. Matching for > EUR 50 (Non-linear Matching): An explanation that an anonymous donor would donate an additional euro for each euro donated, conditional on the donation being greater than EUR 50 (up to a total of EUR 60,000).

  6. Additional EUR 20 for each donation (Leveraged Matching): An explanation that an anonymous donor would contribute an additional EUR 20 for each donation made, regardless of the size of the donation (up to a total of EUR 60,000).

The lead gift was provided by an anonymous donor and each matching scheme was implemented as stated in the letters. The opera did not have a fundraising target in mind, and none of the mailings indicated a specific financial target.

Results and policy lessons

Overall, the mailing campaign elicited 922 individual donations totaling EUR 85,900, EUR 127,039 including the value of matches. Neither the presence of a lead donor nor differences in matching schemes affected individuals’ decisions to donate, but having a lead donor nearly doubled the average amount donated.

Neither the presence of a lead donor nor changes in the amount matched per donation significantly affected individuals’ decision to donate: Response rates varied from 3.5 percent (Lead Donor) to 4.7 percent (Fixed Gift Matching) across treatments, and researchers did not find an impact of the different letters on the number of individual donations.

The impacts on the amount donated varied across treatment groups: Providing information about the presence of a substantial lead donor nearly doubled the size of the average donation relative to the comparison group – from EUR 74.3 to EUR 132. Meanwhile, providing information about the existence of a matching scheme resulted in a decrease in the amounts donated when compared to the lead donor treatment: the average donation fell to EUR 101 in the 50 percent matching, and to EUR 92.3 in the 100 percent Matching. The same decrease was observed in the two remaining intervention groups.

Results suggest that an optimal design for mail solicitations of charitable donations would be one in which the Opera announced the existence of a significant and anonymous lead donor, and did not use the lead donation to match donations in any way.

Huck, Steffen, and Imran Rasul. 2011. "Matched Fundraising: Evidence From a Natural Field Experimen." Journal of Public Economics 95(2011): 351-362. Huck, Steffen, and Imran Rasul. 2014. "Comparing Charitable Fundraising Schemes: Evidence From a Natural Field Experiment and a Structural Model." American Economic Journal: Economic Policy, 7(2):326-69.
1.
Nonprofits Source. n.d. “The Ultimate List of Charitable Giving Statistics for 2018.” Accessed November 26, 2018. https://nonprofitssource.com/online-giving-statistics/.