Home Education The Dynamics of Opportunity in America
Evaluating Resource Levels and Fairness
The next step is to estimate levels of real resources in otherwise comparable settings across states and to estimate variations in real resources with respect to child poverty.
Estimating Staffing Levels and Distributions Our approach to modeling staffing levels follows the one we used to model funding levels. We use annual data from 1993 to 2012 and apply the same model as above, except putting numbers of teachers per 100 pupils on the left-hand side. Again, the premises are: overall staffing ratios might be higher on average (better) in states with more children in small, low-population-density districts; staffing ratios (given spending levels) might be lower (worse) in states facing higher labor costs; and staffing ratios should vary with respect to children's educational needs, as proxied by district poverty measures.
Teachers per 100 Pupils = f (Regional Competitive Wages, District Size *
Population Density, Grade Range Served, State ´ Census Child Poverty Rate)
We then use this model to (a) generate predicted values of teachers per 100 pupils at given levels of poverty, within each state and (b) generate a staffing fairness ratio like our funding fairness ratio.
Evaluating the Average Competitiveness of Teacher Wages As discussed above, one way in which teacher wages matter is that the average relative wage of teachers versus other professions in a given labor market may influence the quality of those entering and staying within the teaching workforce. Here, we use the U.S. Census Bureau's American Community Survey (ACS) annual data from 2000 to 2012 to estimate, for each state, the ratio of the expected income from wages for an elementary or secondary school teacher to the expected income from wages for a nonteacher at the same age and degree level.
Of primary interest here are the differences in competitive wage ratios across states, and ultimately, whether states that allocate more resources to education generally are able to achieve more competitive teacher wages. Here, we compare annual wages of teachers to nonteachers, but we also note that variation across states remains similar with a comparison of weekly or monthly wages, although teacher wages do become more comparable to nonteacher wages. Recall that literature on teacher wages and teacher quality suggests that the more competitive the teacher wage (relative to other career options), the higher the expected quality of entrants to the profession.
To generate our competitive wage ratios, we begin with a regression model fit to our 13-year set of ACS data, in which we estimate the relationship between “income from wages” as the dependent variable, a series of state indicators, and an indicator that the individual is a teacher (occupation) in elementary or secondary education (industry). We include an indicator of the teacher's age and education level, and we include measures of hours worked per week and weeks worked per year but do not equate our predicted wages by holding constant these latter two factors in the analyses. We estimate the following model:
Income from Wages = f (State Place of Work, k12 Teacher, Age, Education Level, Hours per Week, Weeks per Year)
We use this model to generate predicted values for teacher and nonteacher wages at specific age points, for individuals with a bachelor's degree, and then take the ratio of teacher to nonteacher wages. Of particular interest are (a) the differences in the teacher/nonteacher wage ratio across states and (b) the changes over time within states in the teacher/nonteacher wage ratio. That is, are teacher wages more competitive in some states than others? And have teachers generally gained or lost ground? Are these differences in wage competitiveness and gains or losses related back to state funding levels?
|< Prev||CONTENTS||Next >|