The cost of education

Eighty percent of the pie

How does the Hamilton County Department of Education (HCDE) spend its money?

The district is faced with a multitude of expenses, like transporting and feeding students, making sure buildings stay heated and cooled, and supplying textbooks and technology to classrooms. But overwhelmingly, HCDE spends most of its money on people, which translates into salaries and benefits for teachers, counselors, librarians, principals and central office staff. This expense alone takes up more than 80 percent of HCDE’s operating budget.

Putting the ‘fun’ back into funds

We previously broke down HCDE’s different revenue streams. Before we start exploring personnel expenses, we first want to briefly explain how funds get apportioned. All revenue gets broken down into the following four funds:

  • General purpose school fund: Consider this HCDE’s operating budget. Money in this fund is discretionary and goes towards the basic operations of the school system, like paying bills, buying textbooks and paying salaries and benefits. During the 2014–15 school year, HCDE had a budget of $345,319,137 for this fund.
  • Federal projects fund: These funds are earmarked for specific purposes designated by state or federal governments, such as Title I and the Individuals with Disabilities Education Act, which means that not all schools or students qualify. Last year HCDE received $31,679,204 for its federal projects fund.
  • Food service fund: These funds go directly to federal school nutrition programs. The budget for last year was $19,895,927.
  • Self-funded projects fund: Money in this fund comes from grants or donations from non-governmental entities and are usually earmarked for specific schools or programs. The amount of money in this fund totaled $2,276,959 in 2014–15.

Roughly 81 percent of the general purpose school fund goes towards personnel spending — $203 million in salaries and $78 million in benefits in 2015. Since this is such a big slice of the discretionary fund pie, we believe that to understand average per-student expenditure, we first need to understand how much in salaries and benefits is apportioned to each school. HCDE also recognizes the value of breaking down salaries and benefits as a step toward increasing transparency and understanding spending across the county. During its first budget work session on March 10, HCDE Assistant Superintendent of Finance Christie Jordan distributed a school-by-school breakdown of salaries and benefits for each of the nine districts in Hamilton County1. Thanks to these figures we know how teachers and other school staff get compensated for their hard work.

Measuring up

Let’s first look at how teachers are paid across the U.S. and how Tennessee measures up. According to the National Education Association, the largest labor union representing teachers and other school staff, the average salary for a teacher in 2015 was estimated to be $57,379 across the U.S. In Tennessee, the NEA estimates the average salary was $48,503, which was the 39th lowest in the country for 20152. The Tennessee Department of Education reports a slightly lower estimate for the average state salary of $47,979. In contrast, the 3,000 classroom teachers in Hamilton County received an average of $48,298 for the 2014–15 school year, slightly higher or lower than the state average depending on which source you use.

Gov. Bill Haslam has recognized the need to increase teacher pay in Tennessee. He recently committed $104.6 million to salary raises for the 2017 fiscal year. This is part of his larger goal of making Tennessee the fastest improving state for teacher pay in the nation — a goal not yet achieved. NEA statistics for the percent change in teachers’ salaries between the 2012–13 and 2013–14 school years show Vermont was the fastest growing state for salaries with a 4.1 percent increase. The national average increased 1 percent. Tennessee salaries increased 0.4 percent in the same period.

Average salaries of course mask details about how individual teachers are compensated. In Hamilton County, teachers’ salaries are pegged to a pay schedule determined by academic qualification and years of teaching experience. For example, a new teacher with a bachelor’s degree can expect a starting salary of $36,765. If that same teacher continues to work for 25 years at the current pay scale, he or she would eventually earn an annual salary of $54,107.

Breaking it down

We obtained from HCDE a school-by-school breakdown of 2015–16 salaries and benefits for school-level personnel, which includes teachers, principals and all other staff. We used enrollment numbers to determine that the average per-student expenditure on salaries and benefits was $5,633 across 74 schools.

This average increases or decreases depending on whether we separately examine Title I and non-Title I schools. Across Title I schools, the average per-student expenditure jumps to $5,796. For non-Title I schools, the average decreases to $5,398. Fifty percent of schools fall below the countywide average. This includes 46 percent of Title I and 58 percent of non-Title I schools. (We list school-by-school averages in this graphic.)

We can examine school-level salary trends across time in Hamilton County using the U.S. Department of Education’s study on salary spending from the 2008–09 school year. By comparing the percent difference of each school’s average spending to the countywide average for 2008–09 and 2015–16, our analysis highlights which schools got closer to annual average spending and which schools got further away3. Overall, we were able to compare 68 schools between 2009 and 2016. Thirty-three schools got closer to the per-student average across time. Thirty-five got further away. Fifty-five percent of non-Title I schools got closer to the average per-student spending amount in 2016 compared to 46 percent of Title I schools.

Money can’t buy everything

Do different average salaries correspond to different teacher quality and qualification across schools? We wanted to see how our findings on average salaries related to teacher qualification and quality at the school-level. But first we want to acknowledge that measuring teacher effectiveness isn’t easy. Graduation rates, grades and test scores are one way to measure success, but then again, student improvement is just one kind of data point. Teachers can serve a much more expansive role in the way they motivate and inspire students that isn’t always easy to quantify. We also want to point out that our analysis focuses on average school-level data, which cancels out individual variation, both good and bad.


When a school principal recruits a teacher he or she doesn’t have a say in how much that teacher is paid. That’s already determined by the HCDE pay schedule. So schools simply get to pick the best qualified teacher for the job. Different schools, however, have different access to quality teachers. Good teachers may flee or avoid “difficult” schools. This problem has been acknowledged by the U.S. Department of Education, which has called upon states to develop strategies to promote an equitable distribution of excellent teachers across schools. The Tennessee plan, Ensure Equitable Access to Excellent Educators, was released last year.

Tennessee’s plan defines highly effective teachers as those who receive a 4 or 5 on the Tennessee Value Added Assessment System. TVAAS measures how well teachers are improving students’ growth, rather than whether they are proficient on the state assessment. A score of 4 or 5 indicates that students are exceeding expected growth; a score of 3 indicates they are achieving expected growth; and a score of 1 or 2 indicates they are not meeting expected growth. We want to point out that TVAAS scores are only one tool in a much larger kit that helps determine teacher effectiveness. In fact, TVAAS scores usually account for only 35 percent of a teacher’s total evaluation score.

We obtained a sample of anonymous teacher-level TVAAS scores for over 200 teachers in 74 Hamilton County schools. We found that 41 percent of sampled teachers in elementary schools received a 4 or 5 TVAAS score compared to 28 percent in secondary schools. For teachers in Title I schools, these rates are nearly the same. In non-Title I schools, teacher effectiveness is significantly higher. While it appears Title I schools tend to have higher average per-student spending on salary and benefits than non-Title I schools, this does not necessarily translate into better TVAAS scores for individual teachers.

Teachers earning a TVAAS score of 4 or 5
All teachers41%28%
Title I teachers42%26%
Non-Title I teachers52%33%

We also looked at school-level composite TVAAS scores for 77 Hamilton County schools. Trends were somewhat the same as teacher-level TVAAS scores, with elementary schools performing consistently better than secondary schools. Non-Title I schools also had higher rates of 4 and 5 scores than non-Title I schools.

Schools earning a TVAAS score of 4 or 5
All teachers41%15%
Title I teachers37%14%
Non-Title I teachers50%17%


Differences in TVAAS scores between Title I and non-Title I schools may be due to teacher-level differences in academic qualification and experience. Veteran teachers with an advanced degree might be more effective than teachers just out of college, or vice versa. To explore this possibility, we obtained 2014–15 teacher-level data on academic attainment and years of experience for 78 schools in Hamilton County.

We calculated the percentage of first year teachers with bachelor’s degrees and found that across the county around 6 percent of teachers were new to the job. That rate was 4 percent for non-Title I schools and 8 percent for Title I schools. These differences balanced out somewhat when expanding the analysis to teachers with five years experience or less. These teachers represent 19 percent of educators across the county, 15 percent of teachers in non-Title I schools and 22 percent of teachers in Title I schools. Based on this analysis, we saw evidence that teacher experience differs between Title I and non-Title I schools.

We then compared school-level TVAAS scores with data on academic attainment and years of experience for roughly 3,600 teachers at 73 schools. We developed a Teacher Investment Index for each school, which allowed us to score schools by total years of teacher experience and academic attainment4. The results show that there isn’t a strong relationship between our Teacher Investment Index and TVAAS scores. That is, schools with more veteran teachers or advanced degrees, don’t appear to have higher TVAAS scores than schools with less experienced teachers with only bachelor’s degrees.

Our analysis shows that higher spending on salaries and benefits doesn’t necessarily translate into better teachers or student learning. Other factors may play a role in student achievement, such as ongoing teacher training, students’ access to technology and parental support.

What’s next?

Teacher salaries alone do not determine student success. To understand the relationship between average spending and student outcomes, we need to dig deeper into the funding question and add additional factors, like federal and school-level funds. We’ll be exploring these other factors in the weeks ahead.


  1. We applaud HCDE’s step towards greater transparency and clearer budgeting data. However, we think the department can go further. HCDE should publish this salary information on its website so that it’s available to the general public and not just to members of the community who attended the March 10 meeting. If you’re wondering what you missed, we’ve made this data available for download

  2. NEA rankings include Washington, D.C. 

  3. Data for both 2008 and 2016 pertains to all school-level personnel. Data is for salaries only and excludes benefits. 

  4. We want to give two caveats regarding the database used to calculate the Teacher Investment Index. The first is that years of teaching experience can be an imprecise measurement. For example, a new teacher might have a year or more of experience in the classroom as a teaching aide to fulfill an academic requirement, but show zero years of teaching experience as a new hire. Another example is that the database has a ceiling of 25 years, which would not accurately reflect the additional years of experience of someone who has been teaching for 30 years. The second caveat is that the database shows active employees based on pay status. What this means is that if a teacher is on leave of absence but has enough available leave to still be considered in paid status, that teacher will be counted among active teachers. Further, interim teachers are also counted, which can lead to double counting for some positions. 

Jacqueline Homann

Policy research director

Jacqueline Homann has worked at a number of international and federal institutions. She is a graduate of the LBJ School of Public Affairs at the University of Texas at Austin and holds an M.A. in policy studies.