Small Differences that Matter: Mistakes in Applying to College

Higher Education, April 2014

This paper esti­mates the sen­si­tiv­ity of stu­dents’ col­lege appli­ca­tion deci­sions to a small change in the cost of send­ing stan­dard­ized test scores to col­leges. Using con­fi­den­tial ACT micro data, Amanda Pallais finds that when the ACT increased from three to four the num­ber of free score reports that ACT-takers could send, the frac­tion of test-takers send­ing four reports rose sub­stan­tially while the frac­tion send­ing three fell by an off­set­ting amount. Students simul­ta­ne­ously sent their scores to a wider range of colleges.

Using micro data from the American Freshman Survey, two iden­ti­fi­ca­tion strate­gies show that ACT-takers sent more col­lege appli­ca­tions and low-income ACT-takers attended more selec­tive col­leges after the cost change. The first strat­egy com­pares ACT-takers before and after the cost change, con­trol­ling for time trends and covari­ates, and the sec­ond esti­mates difference-in-difference regres­sions using SAT-takers as a con­trol group. Back-of-the-envelope cal­cu­la­tions sug­gest that by induc­ing low-income stu­dents to attend more selec­tive col­leges, the pol­icy change sig­nif­i­cantly increased their expected earn­ings. Because the cost of send­ing an addi­tional (non-free) ACT score was merely $6 through­out, this siz­able behav­ioral change is sur­pris­ing and sug­gests that stu­dents may use sim­ple heuris­tics in mak­ing their appli­ca­tion deci­sions. In such a set­ting, small pol­icy per­tur­ba­tions can have large effects on welfare.