The authors ran a fund-raising field experiment to see the effect of providing social information on contribution rates (n=225). There were three conditions; (i) Donors were told information about another donor’s contribution which was above the original donor’s contribution last year, (ii) donors told information about another donor’s contribution below the original donor’s previous one and (iii) donors told information about another donor’s contribution which matched their own previous one. The messages were in the form “We had another member, they contributed $X. How much would you like to pledge today?”They find that respondents change their contribution in the direction of the social information; increasing their contribution when the social information is above their previous contribution, and decreasing their contribution when the social information is below.