The "Highly Qualified Teacher" requirements of the No Child Left Behind Act (2002) have stimulated increased interest in the level and structure of teacher pay and benefits. National survey data on these variables are available from a variety of sources, including the Decennial Census, the American Community Survey (ACS), the Current Population Survey (CPS), and the Schools and Staffing Survey (SASS). SASS collects data specific to education and was developed by the National Center for Education Statistics (NCES), while the other surveys are conducted by the Census Bureau and collect data about the employment and earnings of household members.
Data collected via these household or teacher surveys generally include information that may be used to control for teacher education and experience; such information is critical when comparing teacher pay. Labor force surveys, such as those conducted by the Census Bureau, also permit teacher and nonteacher pay comparisons. However, these surveys are not without shortcomings. For example, like the Decennial Census and the ACS, the CPS allows for proxy reporting, which may affect the accuracy of reports of earnings. Additionally, income is reported for a 12 month period, making it impossible to separate salary for teaching from salary from teaching or other secondary employment in the summers. Finally, the earnings data may have considerable reporting error and may yield biased estimates because a large and increasing share of respondents fail to report pay.1 In these cases, earnings are imputed. However, when nonteacher earnings are used to impute teachers' earnings, the utility of these data for teacher-nonteacher comparisons is compromised.2 There is no easy fix for nonreporting of earnings in labor force surveys, and the problem is growing.3
Many state education agencies (SEAs)4 collect data about individual school employees. These state administrative data, collected directly from school districts, have the potential to produce more comprehensive and more accurate estimates of teacher pay. They often include variables at the teacher level that reflect teaching assignments, base pay, earnings supplements or extra-duty pay, experience, and demographics. There are clear advantages to utilizing SEA administrative data in analyzing teacher compensation. First, because these data are collected directly from districts (often from district payroll offices), they are more accurate than data collected from household surveys. Second, they are available for the entire universe of public school teachers, not just a sample. This makes possible reliable disaggregated comparisons across subgroups (e.g., districts or teaching fields).
This report explores the feasibility of using teacher-level administrative data maintained by SEAs for interstate or interdistrict studies of teacher pay. In particular, the report attempts to answer the following questions: