In recent years, several states increased the age cutoffs for kindergarten entry requiring that children are at least five years old at the start of their kindergarten school year (e.g. moving the date from December to September). Influencing these decisions is some early evidence of the positive effects from entering school at an older age, and marked increases in educational accountability and performance demands for secondary and elementary education; these factors have increased the importance of policy decisions concerning early education (Datar, 2006).  Yet the evidence on whether age at school entry is beneficial for children’s outcomes in the short- and long-term is equivocal.  

            In 2009, North Carolina moved the kindergarten birthdate cutoff from October 15^th to September 1^st, requiring kindergarten entrants to be 5 years old on or before September 1^st of the 2009-10 school year (and thereafter). We use recent (2007-2013) statewide micro-level census data from North Carolina, including student’s exact birthdates and information from kindergarten through 3rd grade, to examine whether this policy change influenced student achievement on reading and mathematics exams at the end of 3^rd grade. We test whether this change in the cohort age distribution influenced the statewide test score distribution. To control for school district-level influences, we conduct both statewide and within-district difference-in-difference models. For robustness, we also compare the test score gap between 3^rd and 4^th graders who entered school under the October 15^th cutoff policy with the test score gap between 3^rd and 4^th graders who entered school under the September 1^st cutoff policy. In addition, we examine changes in the proficiency gap between 3^rd and 4^thgrade students under the two cutoff policies.

            We also examine whether this policy change affected the incidence of kindergarten ‘redshirting’ - students enrolling in kindergarten a year later than as prescribed by state birthdate cutoffs. In a prior study of the effects of kindergarten redshirting in North Carolina, we found that redshirted students were overwhelmingly more likely to be designated by their school as having a disability - up to 2.8 times the risk of being designated as disabled in 3^rd grade. Therefore we also test whether the birthdate policy change influenced other available outcomes by the end of the third grade year (i.e. designation as gifted, retention before 3^rdgrade).

            We are in receipt of test score and demographic data from the 2012-13 school year capturing the first year of statewide testing data for kindergarten students enrolled under the new cutoff policy and our preliminary analyses show changes in the test score gap and proficiency rates across grade levels. The statewide incidence of redshirting decreased by 1.8 percentage points, going from 4.3% in 2008 to 2.6% in 2009, and dropping to 2.2% in 2010.

            This is the first statewide evaluation of a change in birthdate cutoff policy in a large and diverse state. More and more states are considering policy changes to birthdate cutoffs; the findings from our study are therefore extremely relevant in the current policy education policy discussion across the country.