This experimental study tests which of the following two explanations of why people vote for charitable donations is behaviorally more relevant. According to the low-cost theory, people trade-off a "warm glow" from the act of voting for charitable donations against the cost of voting for donations. These costs depend on whether the voter expects to affect the material outcome of the ballot. According to conditional cooperation theory, people are willing to vote for charitable donations if they expect many others to vote for donations. The results show that the low-cost theory clearly fails to explain individual and overall voting behavior. It is shown that subjects tend to vote for charitable donations if they expect many others to vote for the donation.