Abstract:
In this paper, we present evidence from a large-scale natural experiment in which different fundraising schemes are tested and compared: lead donor, standard linear matching (1:1), and an alternative matching scheme in which the matching amount goes to another project. The lead donor treatment, in comparison to other treatments, generates high average donations but relatively low participation rate. In the case of the standard linear matching scheme (1:1) we find, similarly to Huck and Rasul (2011), a crowding-out effect, which reduces significantly the average donation given. However, in our case, the traditional linear matching scheme (1:1) increases significantly the participation rate such that the overall return per mail-out is also higher. The alternative matching scheme increases significantly the response rate without generating the above crowding out effect.