Charitable deductions: modelling of longitudinal taxpayers’ data for time-regularity estimation and future predictions

Abstract

Abstract The present study is the continuation of an existing one, with a new research question and novel statistical predictions. It represents the first large-scale time-series analysis of tax incentives for charitable giving in Switzerland for regular donors, and one of the few charitable deductions’ projections globally. Using unique panel data of the entire Geneva taxpayers’ population, from 2001 to 2011, we concentrate on the frequency of donations, by analyzing the characteristics of regular donors. This period also encompassed a legal reform that raised ceilings for charitable deductions. Taxpayers who deducted only once represent 29% of the study population, while the remaining 71% indicate taxpayers deducting more than once. We also observe that regular donors are not driven by the ceilings for tax incentives, however, wealth is an important variable characterizing the donations’ frequency. Moreover, time-series forecasting models have been used to project the future amount of donations and donors. A total amount of donations in a range of (79, 114) million in 2012-2022 has been predicted from the best model. These figures help understanding the magnitude of deductions and they allow to take better decision of the ceiling of tax incentives for charitable giving, while highlighting the main relevant factors.