The impact of education reforms on salaries of Arizona school district superintendents
Objectives: To determine whether superintendent salaries for 86 Arizona school districts during the 1995-2005 time period were influenced by [a] district performance on two formal accountability systems (No Child Left Behind and AZLearns, the latter system sponsored by the state of Arizona) and [b] charter school market share at the district level.
Data Sources: Annual superintendent salary data for 1995-2005 for all Arizona school districts with elementary schools (elementary or unified; high school districts were excluded); with an enrollment of at least 1000 students in the 1994-95 school year; and which were not located on Indian Reservations, was collected from various Annual Reports of the Arizona Superintendent of Public Instruction. Data on district characteristics were obtained from the Common Core of Data and the School District Demographics System (U.S. Dept. of Ed.). Accountability performance data for K-8 schools for 2002-2004 (when NCLB and AZLearns were initiated) was obtained from the Arizona Dept. of Education. District market share data for charters (which started in 1995) was obtained for Fall 1997 and Fall 2003 from prior studies.
Methods: [1] OLS regressions, with annual salary as the dependent variable, for six years (1997-99 and 2003-05), and [2] First-difference regressions for the 1995-2000 and 2000-05 time periods for superintendents in place during those time periods.
Results: OLS regression results find that both district enrollment size and (to a lesser extent) a measure of labor market competition for superintendents had a positive impact on salaries for both 1997-99 and 2003-05; in addition, higher performance on AZLearns (but not NCLB) may have increased salaries during 2003-05. Charter market share did not influence salaries in the OLS regressions. In contrast, first-difference results for 2000-05 find that large increases in charter market share from 1997 to 2003 reduced salaries, while higher AZLearns performance boosted salaries. As with OLS, NCLB did not influence salaries in the first-difference regressions. The first-difference results are tentative due to small sample sizes.
Tentative Conclusions: All else equal, Arizona school districts whose elementary schools performed well on AZLearns, and who did not experience a large increase in charter school competition from 1997 to 2003, paid their superintendents more during the 2000-2005 time period. Possibly, these salary increases at the senior leadership level may have been due to increased district school effectiveness.