Why There Are Not Enough STEM Workers in the U.S. Labor Market

shutterstock_85654096Occasional research, such as a report released last week by the Economic Policy Institute, suggests the U.S. has a sufficient supply of science, technology, engineering, and mathematics (STEM) graduates and workers. However, these conclusions are at odds with a growing number of expert analyses that find the U.S. does in fact face significant challenges in meeting the growing needs of our expanding knowledge-based economy.  Here is a sampling of the evidence:

STEM degree holders have very low unemployment rates and use their skills in STEM and non-STEM occupations. One of the arguments frequently used to show the absence of shortage is that not all STEM graduates are hired into STEM jobs.  However, there is strong evidence that these degree holders are doing pretty well in finding jobs, even if those jobs are outside of STEM fields. The unemployment rates for individuals with advanced STEM degrees are very low: 3.1 percent for U.S citizens with STEM PhDs and 3.4 percent for those with STEM masters degrees, and often much lower for specific occupations. The fact that individuals with STEM expertise are working in non-STEM designated occupations doesn’t mean that they have been pushed out of their field by competition with foreign workers. Similarly, the fact that some STEM degree holders say they are finding better job opportunities outside of IT occupations doesn’t mean that IT opportunities are bad.  It may simply mean that opportunities in other areas are better and that more companies are competing for this relatively scarce talent.  Specifically, research suggests as many more companies seek to improve innovation or expand technology use, there is a broader competition for STEM competencies – knowledge, skills and abilities such as research, analysis, quantitative and computer skills, and methodologies – among many businesses and occupations far beyond specific STEM-defined fields.

STEM employers report thousands of unfilled positions. Due to geographic spatial and skills mismatches, employers report difficulty recruiting individuals with specific skills, particularly in metropolitan areas where innovation industries agglomerate. In 2010, for example, San Francisco and San Jose had 25 and 19 job openings for every computer graduate, respectively, and in Austin, Des Moines, Washington, Charleston, Seattle, and Charlotte, that gap was nearly as great. Moreover, a 2011 report describes 3,200 unfilled positions at Siemens, and Microsoft reported over 6,000 unfilled job openings in the United States in 2012, over half of which are for researchers, developers, and engineers. Collectively, according to the Silicon Valley Leadership Group, IBM, Intel, Microsoft, and Oracle have 10,000 U.S. job openings. If there is no STEM labor shortage, why are so many good paying innovation industry jobs going unfilled?

Growing evidence shows that the H-1B visa program has a positive effect on wages of U.S. workers, and local economies benefit from innovation industry jobs with strong multiplier effects. A 2013 report of 219 U.S. cities from 1990 to 2010 shows that H-1B driven increases in STEM workers in a city were linked with statistically significant increases in wages for all college-educated native-born workers. Additionally, a one percent increase in the foreign-born STEM worker share of total employment in a city increased the wages of both STEM and non-STEM college-educated workers by 4 to 6 percent. Furthermore, innovation industry and STEM jobs tend to have higher than average multiplier effects. Such effects allow for greater job creation in metropolitan areas with strong innovation industry economies. Indeed, an analysis of 320 of the 381 metropolitan statistical areas in the United States shows that each new high-tech job in a metro area creates five additional long-term local jobs outside of the high-tech sector across the skills spectrum. In many U.S. metropolitan areas, the innovation economy, and the high-skilled jobs related to it, drive prosperity for a broader base of workers living in the region.

In summary, a more precise understanding of the visa programs and employment data is needed for a balanced understanding of the overall labor market for STEM skills in the United States. As the innovation industry is critical to economic growth locally and nationally, ensuring a robust labor force with a strong STEM skills set is important now and in the future.

Tags: , , , , ,


If you enjoyed this article, subscribe now to receive more just like it.

Subscribe via RSS Feed
  • EPI Report disputes your analysis.

    Key Points from the EPI Report:

    -Guest workers are filling as many as half of all new IT jobs each year

    -IT workers earn the same today as they did 14 years ago

    -Currently, only one of every two STEM college graduates is hired into a STEM job each year

    -Policies that expand the supply of guestworkers will discourage U.S. students from going into STEM, and into IT in particular

    • Paul McDaniel

      Correlation is not causation, and EPI should demonstrate how immigration has anything to do with their assertion that only half of STEM graduates end up in STEM-defined positions, or that those graduates aren’t finding good jobs outside STEM. Meanwhile, the studies and research centers I linked to in the blog post suggest otherwise – that those with STEM skills and competencies are reporting that they’re using those skills in a wide range of jobs, not just jobs that are specifically defined as STEM occupations.

      • Hi Paul –
        I’m concerned about the identification of the models that produce those enormous multiplier estimates. They seem implausible on their face, given what we know about employment multipliers.

        Do you really think instrumenting by local immigrant population (the Zavodny study, for example) gets at causality? I find that very hard to swallow. These large multiplier estimates are most likely coming from the fact that demand for immigrants is strongest in growing areas.

        In other words – I think it’s your claims here that need to be more careful about causality. But if you think it’s a good instrument I’m happy to hear the case.

        I don’t think it’s strong enough to discriminate between low and high skill immigrants and say that high skill immigrants are more desirable to have here than any other kind.

        • Hi Daniel,

          Thank you for your comments.

          I never said that high-skill immigrants are more desirable than less-skilled immigrants, nor would I make that argument. All immigrants are desirable and make important contributions. Immigration overall has a positive benefit nationally and locally as documented throughout the literature (which we review in a variety of resources at http://www.immigrationpolicy.org/just-facts/economics-immigration-resource-page). It’s not just that demand for immigrants is strongest in growing areas, but demand for everyone is strongest. Growing areas are just that: growing. There is a demand for all skills. Metropolitan areas with high agglomeration of innovation industry activity attract high-skilled workers. Those jobs in turn create additional growth opportunities in other industry sectors across the skills spectrum. The demand generated by metropolitan regional economies is why, in addition to national level analysis, we must also look at metropolitan level.

  • I would like the author to provide some hard facts to support these assertions:

    “The fact that individuals with STEM expertise are working in non-STEM designated occupations doesn’t mean that they have been pushed out of their field by competition with foreign workers. Similarly, the fact that some STEM degree holders say they are finding better job opportunities outside of IT occupations doesn’t mean that IT opportunities are bad.”

    • Paul McDaniel

      I believe I linked to those studies in the blog post. But here they are again:
      The Center on Education and the Workforce at Georgetown University – http://cew.georgetown.edu/stem/ – finds that there is a broader competition for STEM competencies among a wider array of businesses and occupations far beyond the narrowly defined band of STEM fields. Labor market demand for STEM competencies beyond specific STEM-designated occupations suggests a potential shortage of workers with skills sets to fill positions with STEM and non-STEM employers and occupations. Moving forward, we will need a healthy supply of workers with STEM competencies although not exclusively traditional STEM workers working in traditional STEM jobs. Indeed, as an example, a 2012 National Science Foundation (NSF) study – http://www.nsf.gov/statistics/seind12/pdf/seind12.pdf – projects that all science and engineering (S&E) employment will increase by greater than 20 percent through 2018. In particular, computer and mathematical occupations are expected to grow by more than 25 percent. In addition to projections, the NSF report examines the relationships between degrees and jobs, and the associated competition for competencies. Specifically, the report describes the perspective that individuals with highest S&E degrees working in S&E-related and non-S&E occupations still view themselves as working within their degree field. This means that even though an advanced S&E degree holder is not working in a specific STEM-defined occupation, he or she still identifies the job as one in which S&E skills are used.

      • Concerned IT Worker

        So you are making the case that the current level of H1B workers are suppressing the wages in the STEM fields? If STEM graduates would rather work in what they aren’t trained for, than taking a STEM job for low pay, then I see that as a big problem. Down the line US students won’t bother to pursue these fields of study.

  • I think what some of your commenters are skeptical about is that you ignore the decades of careful economic research that suggests there isn’t really a problem with shortages (Blank and Stigler, Freeman, Stephan, Ryoo and Rosen, forthcoming Salzman, Kuehn, and Lowell, Arrow and Capron, Hansen, Butz et al., Barnow et al., etc.) and instead just toss out widely panned reports like Carnevale’s.

    It gives the impression that either you are cherry-picking or you’re not familiar with the literature.