A field experiment in Minnesota in 1994 tested the efficacy of normative messages in letters to taxpayers. There were 3 conditions in total; 2 treatments and a control, each consisting of 20,000 people. The first treatment group received the standard tax letter plus a rational appeal for paying their taxes (“your taxes fund local services such as X, Y, Z”). The second treatment received the standard letter plus a social normative message (“People who file tax reports report correctly and pay voluntarily 93% of the income taxes they owe. Although some taxpayers owe money because of minor errors, a small number of taxpayers who cheat owe the bulk of unpaid taxes”).The results found no significant effect of the normative messages on tax declarations.