Table of Contents
Chapter 1 The Long View: Trends in Teacher Pay
Chapter 2 Recent Trends in the Relative Earnings of Teachers
Chapter 3 The Earnings of Teachers Relative to “Comparable” Occupations
Chapter 4 The Role of Differences in Non-Wage Benefits
Chapter 5 Misuse of the National Compensation Survey, Revisited
Conclusion and Policy Implications
About the authors
Teacher quality is the most important input schools contribute to the academic success of their students (Hanushek and Rivkin 2006; Rice 2003). Yet for many school officials, recruiting and retaining talented and effective classroom teachers remains an uphill battle. For decades now, a small and declining fraction of the most cognitively skilled graduates choose to become teachers (Corcoran, Evans, and Schwab 2004), while rigorous national standards and school-based accountability for student performance have pushed the demand for talented teachers to an all-time high.
Recent efforts to recruit and retain highly skilled teachers have reenergized the debate over teacher compensation. Many continue to ask whether teacher salaries are sufficient to attract the best graduates into teaching (Stronge, Gareis, and Little 2006; Moulthrop, Calegari, and Eggers 2005), while others minimize the importance of base pay and question whether the current structure of teacher compensation is optimal for attracting talent into the profession (Solmon and Podgursky 2000; Hoxby and Leigh 2004; Leigh and Mead 2005). Whatever the case, it is clear that sound evidence on the comparability of teacher pay remains critical to our understanding of the link between compensation and teacher quality and those policies that will ensure a cadre of teachers capable of helping students to meet increasingly higher achievement standards.
In an earlier study (Allegretto, Corcoran, and Mishel 2004) we contributed to this evidence by examining recent trends in the relative weekly earnings of elementary and secondary school teachers. In that study we found that the average weekly pay of teachers in 2003 was nearly 14% below that of workers with similar education and work experience (p. 13), a gap only minimally offset by higher non-wage benefits in teaching. Though teacher earnings have fallen below that of the average college graduate in recent decades, we showed that teachers lost considerable ground during the late 1990s, as earnings of college graduates grew 11% relative to 0.8% growth in teaching.
In this new study, we extend our earlier work by updating this analysis of relative teacher earnings through 2006, further disaggregating these t rends by seniority level. Using decennial Census data, we place the recent trend in relative pay into its long-run context, and consider how long-run trends in teacher pay complicate efforts to maintain a constant level of teacher quality.
In our 2004 study, we highlighted some important methodological issues that frequently arise in the analysis of teacher compensation, and we revisit some of those issues here. For example, we explain how researchers can sometimes arrive at contradictory conclusions about relative teacher pay when relying on different sources of data. These differences depend on (1) whether the analyst uses an employer- or employee-based survey of earnings, and (2) the pay interval (annual, weekly, or hourly) the analyst elects to compare. Often in hourly wage comparisons (Greene and Winters 2007, for example), quite unrealistic assumptions are made regarding teacher work schedules. All of the data available show that teachers work at least as many hours each work week as comparable college graduates.
The major findings of this report are as follows:
• An analysis of trends in weekly earnings shows that public school teachers in 2006 earned 15% lower weekly earnings than comparable workers, a gap 1 percentage point larger than that reported for 2003 in our original study. The teacher disadvantage in weekly earnings relative to comparable workers grew by 13.4 percentage points between 1979 and 2006, with most of the erosion (9.0 percentage points) occurring in the last 10 years (between 1996 and 2006).
• Recent trends represent only a small part of a long-run decline in the relative pay of teachers. Using U.S. Census data we show that the pay gap between female public school teachers and comparably educated women—for whom the labor market dramatically changed over the 1960-2000 period—grew by nearly 28 percentage points, from a relative wage advantage of 14.7% in 1960, to a pay disadvantage of 13.2% in 2000. Among all public school teachers the relative wage disadvantage grew almost 20 percentage points over the 1960-2000 period.
• An analysis of the weekly earnings of occupations comparable to K-12 teachers confirms the teacher disadvantage in weekly earnings and the substantial erosion of teacher relative pay over the last 10 years. Teachers’ weekly wages were nearly on par with those paid in comparable occupations in 1996 but are now 14.3%, or $154, below that of comparable occupations.
• Improvements in the non-wage benefits of K-12 teachers partially offsets these wage differences, such that the weekly compensation disadvantage facing teachers in 2006 is about 12%, about 3 percentage points less than the 15% weekly wage disadvantage.
• After disaggregating trends in relative compensation through the 1990s by age, nearly all of the increase in the weekly earnings gap between teachers and comparably educated and experienced workers occurred among mid- and senior-level teachers. Early-career teachers (age 25-34) experience roughly the same wage disadvantage today as in 1990 (about 12%).
• Raising teacher compensation is a critical component in any strategy to recruit and retain a higher quality teacher workforce if the goal is to affect the broad array of teachers—that is, move the quality of the median teacher. Policies that solely focus on changing the composition of the current compensation levels, such as merit or pay-for-performance schemes, are unlikely to be effective unless they also correct the teacher compensation disadvantage in the labor market.
• A broad array of analysts from across the political spectrum have found trends comparable to ours—that teachers face an earnings disadvantage, and that this disadvantage has grown over the long run. Only two widely cited analysts seem to disagree with this finding, but the data they examine are inappropriate for this task, as the Bureau of Labor Statistics clearly warns in a statement on its Web site.
• States vary widely in the extent to which public school teachers are paid less than other college graduates. In 15 states, public school teacher weekly wages lag by more than 25%. In contrast, there are only five states where teacher weekly wages are less than 10% behind, and no state where teacher pay is equal or better than that of other college graduates.