State education officials have a new tool to help them predict which investments will pay dividends as they try to boost the number of college graduates who major in science, technology, engineering, and mathematics (STEM) fields.

Defense company Raytheon introduced an open-source program July 8 that will be customizable for the country’s largest school districts, colleges, and state education systems. The program, called the U.S. STEM Educational Model and supported by the Bill & Melinda Gates Foundation, will help officials analyze how they should allocate budgets that have stagnated or shrunk during the economic recession as they seek to increase the number of STEM-related graduates. The program is available for download free of charge, according to the download site.

The computer-based model will simulate how schools can draw students to STEM fields most effectively–a trend that would bolster the science and engineering workforce.

“It can show a return on investment,” said Brian Fitzgerald, executive director of the Business-Higher Education Forum, an organization that hosts the Raytheon modeling tool. “It can show you in a very dynamic way what the effects of alternative investments might be. … This is a very, very good tool for state-level policy makers who are looking for the most effective investment.”

The program can test more than 200 variables that could better inform policy makers about how programs should be funded. The model measures graduation and dropout rates, gender gaps in STEM fields, teacher and STEM industry salaries, and educator attrition rates, among other factors. Raytheon has used a similar program designed to project outcomes of its weapons production for the U.S. military.

The model recently predicted the long-term impact of California’s decision to create smaller classes by hiring more teachers. Many of the new teachers were underqualified, and California decision makers ultimately abandoned the initiative. Fitzgerald said the new computer simulation was fed information about the California program, and within seconds, it showed the possibility of the initiative’s demise.

“It accurately captured those effects,” he said, adding that the computer-based model will change as the open-source community adds new components. New research and data will be included in the newest versions of the simulation model.