We know that different school districts spend vastly different amounts per pupil—anywhere from $8,400 in Charlotte to over $20,000 in Newark. We also know that funding varies across schools within a district. But it's typically difficult to calculate how much each district spends on each individual student—whether Daniella gets more resources than David, or certain groups get more than others.
However, ERS has collected an extensive database of student-level data through our nearly 10 years of working closely with some of the largest urban districts in America. This includes information on teacher salaries, course schedules, and class sizes. Recently, education researcher Rebecca Wolf analyzed data from one large urban school district (with more than 40,000 students) to see how resources were allocated to individual students. She found three main things:
These findings do not mean that the district should necessarily move all extra dollars from high-achieving students to low-achieving students; there are often good reasons to offer smaller and more advanced classes with more experienced teachers to exceptional students. On the other hand, if the district’s main goal is to advance student achievement across the board, particularly for students who are often underserved, it does highlight the need to ensure students receive the resources they deserve. A few of Wolf’s suggestions:
At ERS, we encourage districts to explore alternatives for giving students access to smaller, advanced, or specialized courses. Many can be offered in more cost-effective ways—perhaps through partnerships with local non-profits (like arts organizations or the YMCA), through technology and distance learning, or through sharing teachers across schools.
We are very excited that Wolf has been able to make use of our student-level databases in this way, and we welcome working with other researchers. Contact us if you have a research project you'd like to pursue.
Want to talk to someone directly?
Send us an email at: firstname.lastname@example.org