Measuring High School Graduation Rates at the State Level: What Difference Does Methodology Make?

John R. Warren, University of Minnesota
Andrew Halpern-Manners, University of Minnesota

Warren (2005) developed a measure of state-level high school completion rates using data from the Common Core of Data (CCD). In contrast to other CCD-based measures of states’ high school completion rates, Warren’s (2005) measure is systematic biased by neither 9th grade retention nor migration. We will extend Warren’s (2005) work by providing more current state-level estimates (for 2004); by producing separate sex- and race/ethnic-group specific state-level estimates; and by carefully accounting for international migration. We will also investigate the consequences for the results of substantive analyses of using alternate CDD-based measures. For example, we will ask whether estimates of the effects of state education policies, states’ economic circumstances, or other factors on states’ high school completion rates are dependent on how states’ graduation rates are measured.