Abstract
There is some evidence from field studies and natural experiments that levels of charitable donation depend on the method in which donations are solicited. There is also some experimental evidence that spending on private consumption depends on how much loose change people have. We use a simple laboratory experiment to measure the effect on donor choices of (i) whether the choices are presented verbally or non-verbally, and (ii) whether the participants have a large amount of loose change. We find strong evidence for both effects. These effects may explain some of the variation in the average level of generosity found in different Dictator Game results, and why laboratory experiments elicit levels of generosity that are often much higher than in non-laboratory settings.