Association for Public Policy Analysis & Management

Giving Tuesday, a social-media event encouraging giving on the Tuesday after Thanksgiving, is becoming an important part of year-end giving in the United States. Giving Tuesday (GT) has grown quickly since its inception in 2012, with the 2018 drive raising approximately $380 million. Community campaigns, which typically combine training, marketing, and contests with prizes, have become a substantial vehicle for publicizing Giving Tuesday and encouraging nonprofits to participate. However, some observers are concerned that GT community campaigns may displace, rather than increase, overall year-end giving.

Although peer-reviewed research regarding Giving Tuesday is sparse, research on crowding-out of donations across organizations and time has increased in the past 5 years. Non-experimental studies (e.g. Woronkowicz 2018) have focused on the effects of capital campaigns on the ecology of the local nonprofit sector. Although results from lab experiments are mixed, it does seem that a crowding-out effect exists, at least when the between asks is short (Ek 2017) as opposed to long (Scharf et al. 2018).

This paper uses transaction-level giving data from across the country and data on the locations and dates of community campaigns to estimate a differences-in-differences model. Data on the time and location of community campaigns was developed by combining lists from Giving Tuesday administrators with data collected from websites. Our gift transaction database contains approximately 10 years of anonymized individual donation data from 11,000 nonprofits, so researchers can observe how donors shifted the dates and amounts of their giving over time. The transaction data also includes the location, although not the name of, the nonprofit organizations, allowing us to study the effects of location-specific shocks on giving.

The effect of Giving Tuesday on year-end-giving is identified using a difference-in-differences analysis. Because community campaigns were implemented over time, a standard difference-in-differences analysis or event study is inappropriate and biased (Goodman-Bacon 2018). To overcome this bias, we compute event-by-event effects, which allow us to observe heterogeneous effects between earlier and later adopters of Giving Tuesday community campaigns. In addition, because our data is available at the individual level by organization, we are able to identify the overall effect of Giving Tuesday community campaigns, as well as the specific effects (treatment-on-the-treated, or TOT, effects) on treated organizations and donors. These results contribute to the overall discussion about the effects of Giving Tuesday and also address larger questions related to donor loyalty to organizations over time.

Bibliography

Ek, C., 2017. Some causes are more equal than others? The effect of similarity on substitution in charitable giving. Journal of Economic Behavior and Organization, 136, 45-62.

Goodman-Bacon, A. 2018. “Difference-in-differences with variation in treatment timing.” Working paper No. w25018, National Bureau of Economic Research. Available at https://www.nber.org/papers/w25018.

Scharf, K., S. Smith, and M. Wilhelm, 2017. “Lift and Shift: The Effect of Fundraising Interventions in Charity Space and Time.” CEPR Working Paper. Available at: https://cepr.org/active/publications/discussion_papers/dp.php?dpno=12338

Woronkowicz, J. 2018. “The Effects of Capital Campaigns on Local Nonprofit Ecologies.” Nonprofit and Voluntary Sector Quarterly, 47(3), 645-656.