CORD Lecture Series: Alasdair Rutherford
Regression Discontinuity and Administrative Data: Using Causal Methods to Evaluate Public Policy in the Regulation of Fundraising
Speaker: Alasdair Rutherford is professor of social statistics, University of Stirling, Scotland.
Abstract: The availability of large-scale administrative data on charities and non-profits have provided many opportunities to both test theories and evaluate public policy. Causal methods (e.g. regression discontinuity; interrupted time series; natural experiments) provide opportunities to make causal estimates of interventions when the design of the policy includes (either intentionally or unintentionally) a feature which effectively randomizes organizations into treatment and control groups. In this seminar, Rutherford will discuss the use, opportunities and challenges of regression discontinuity designs with administrative non-profit data, using the example of evaluating participation by U.K. charities in the program of a new Fundraising Regulator. By linking data from different regulators, we are able to exploit a sharp eligibility threshold by fundraising expenditure in order to make a causal estimate of the effect of a “fundraising levy” on participation in the regulation, particularly amongst smaller fundraising charities.