The Cost of Removing Optional Practical Training for STEM Graduates

Workers who possess training or skills in science, technology, engineering, or math (STEM) are some of the most sought after segments of the labor market. Chronic shortages of specialized, high-skilled workers have been making headlines in the United States for nearly a decade, even during the Great Recession. To attempt to address this, in 2016, the Department of Homeland Security revised the extensions to practical training available to international students who had graduated from U.S. institutions with degrees in approved STEM fields. Rather than the normal 12-month period of temporary work authorization allowed regular international graduates from U.S. schools, STEM graduates were allowed an additional 24 months. Given that in 2016, there were 13 times more STEM jobs posted online than there were available unemployed STEM workers in the United States, allowing most U.S.-educated international STEM graduates the opportunity to use their skills to benefit American businesses and workers was one significant step forward in addressing this crisis.

However, despite continuing shortfalls of STEM labor due to tight labor markets and an expanding economy, the current administration has made clear its intent to remove the STEM OPT extension. This change in policy has the potential to cut short the new careers of many STEM workers and to disrupt the production and work of their U.S. employers—both factors sure to have a significant economic impact. To gauge how significant this impact could be, NAE examined data from the 2016 American Community Survey (ACS) and the 2015 National Survey of College Graduates (NSCG) to estimate what the economic impact of rolling back the STEM OPT extension would be.

Our estimates reveal that such action would be extremely costly for the United States. In just one year, the direct costs—the immediate loss of wages and the cost to rehire employees—would total more than $130 million dollars. This is due largely to the relatively high wages and in-demand skills of recent STEM graduates. 

Also, contrary to critics of high-skilled immigration, we find that the vast majority of STEM degree holders who do not work in STEM jobs report that they chose to do so not because they did not think that the pay was high enough, but because of family or personal reasons, or reasons related to their working conditions. STEM degree holders who switch to non-STEM jobs such those in financial services for pay- or promotion-related reasons are paid approximately the same as their STEM-employed counterparts.

Lastly, we find that STEM occupations take longer time to fill than the U.S. average, indicating an acute shortage in STEM workers regardless of nativity. Removing the STEM OPT extension would immediately worsen these shortages and put even more pressure on the already oversubscribed H-1B program in addition to the direct economic costs.

The annual cost of removing STEM OPT would be $130,661,096

  • The total direct cost is calculated by taking the number of STEM OPT recipients in 2014 and multiplying it by the average wage of foreign-born STEM workers between the ages of 21 to 28, multiplied by the time it takes to fill a job (for more on this, see below).
  • Importantly, this number does not account for the full extent of the potential loss. For example, we do not account for the loss of any patents that would have been created by highly skilled foreign-born workers who may no longer be able to work in the United States after finishing their degrees.

Recent STEM graduates, regardless of nativity, earn significantly more than non-STEM graduates, adding important economic stimulus to the national and local economy.

  • Recent STEM graduates[1] earn on average $56,314 annually compared to $35,280 for non-STEM employees in 2015.[2] After adjusting for age, sex, race, and ethnicity, the wage benefit for being a STEM graduate is an average of $18,495.

STEM degree holders are 12.6 percent less likely to be unemployed than non-STEM degree holders.

  • Even with extremely low unemployment rates for college-educated workers, STEM degree holders are 37 percent less likely to be unemployed as recent non-STEM graduates. After controlling for all other factors, this means STEM-degree holders are about 12.6 percent less likely to be unemployed.

STEM degree holders are less likely to be mismatched, that is, to work non-STEM jobs.[3] We find that only those working non-STEM jobs because of personal, family, or working condition reasons face a salary penalty.  

  • STEM degree holders are 16.4 percent less likely to be mismatched than the non-STEM college-educated U.S.-born population.
  • STEM degree holders who work non-STEM jobs because of pay or promotion-related reasons earn approximately the same amount as STEM degree holders who work in STEM. For those mismatched for family, personal, or working condition reasons, we find that they earn anywhere from 44 to 53 percent less than their counterparts in STEM fields or non-STEM jobs held because of pay or promotion-related reasons.

STEM occupations generally take longer to fill than the average time it takes to fill a job in the United States, 41 days. This, on top of the fact that there 16 STEM job openings posted online for every unemployed STEM worker,[4] indicates a severe shortage in STEM-skilled workers regardless of nativity.

  • Looking at the top 30 occupations with highest number of H-1B requests, more than two-thirds of them are STEM occupations.
  • Using the data from Burning Glass Technologies, we find that it takes longer than average to fill the above STEM jobs.
  • For example:
    • Computer systems analyst positions take an average of 56 days to fill—36 percent longer than the average time it takes to fill a job in the United States.
    • IT project manager and technology consultant positions take 59 days to fill. This is more than 40 percent longer than the U.S. average.

[1] Defined as those who are between 21 and 28 years old, have at least a bachelor’s degree, have positive earnings, are not self-employed, and are not enrolled in school.

[2] NAE Analysis of microdata from 2015 American Community Survey, downloaded from

[3] NAE Analysis of data from 2015 National Survey of College Graduates

[4] New American Economy, “Sizing Up the Gap in our Supply of STEM Workers: Examining Job Postings and Unemployment Data from 2010­–2016,” March 29, 2017. Available at:

About Us

New American Economy is a bipartisan research and advocacy organization fighting for smart federal, state, and local immigration policies that help grow our economy and create jobs for all Americans. More…