Sections

Research

Improving workforce development and STEM education to preserve America’s innovation edge

Darrell M. West
Darrell M. West Senior Fellow - Center for Technology Innovation, Douglas Dillon Chair in Governmental Studies

July 26, 2023


  • With AI becoming more ubiquitous in almost every sector, it is outpacing the current skills of workers and requiring considerable adaptation on the part of industrial-era employees.
  • Currently, 45% of STEM employees in the U.S. with a doctoral degree are foreign-born, and the backlash against immigration could pose barriers to retaining and growing that talent pool.
  • The U.S. must prioritize effective implementation of the CHIPS and Science Act, boosting domestic manufacturing capabilities, better workforce development and STEM education, lifelong learning, reasonable immigration policies, and effective management of geopolitical tensions.
A worker miniature is placed among printed circuit boards with semiconductor chips in this illustration picture. July 5, 2023. REUTERS/Florence Lo/Illustration/File Photo.
Executive summary

There are many challenges facing the United States as it shifts from an industrial to a digital economy, but one of the most fundamental problems lies in regard to workforce development and Science, Technology, Engineering, and Math (STEM) education. New advances in robotics, artificial intelligence, and advanced manufacturing are automating jobs, displacing workers, and requiring new skills from Americans. Computer chips are crucial to many different sectors and the construction of new fabs already is facing employee shortages. With skills gaps in several areas, there could be constraints in coming years on the country’s ability fully to reap the advantages of a digital economy.

In this paper, I argue more effective STEM education and workforce development are necessary for U.S. firms to have the people they need for domestic production, supply chain management, cybersecurity, data analytics, and innovation in general. The shift to a digital economy requires bold actions by government, business, and higher education. Failure to make necessary changes will slow the transformation and impede the country’s ability to benefit from emerging opportunities.

Among the actions that will be required include effective implementation of the U.S. CHIPS and Science Act, boosting domestic manufacturing capabilities, improved workforce development training, and more broadly, STEM education, as well as the prioritization of lifelong learning, reasonable immigration policies, and effective management of geopolitical tensions.

Current challenges

Virtually every day, new announcements are made about advances in generative AI, robotics, and advanced manufacturing. Generative AI employs large language models to produce text, images, videos, software, and many other things. Robotics is transforming manufacturing and warehouse operations, and new technologies are being incorporated in factories. Between these things and new developments in quantum computing, biotech, and cloud computing, technology innovation is racing ahead and posing ramifications for many aspects of human life.

With AI becoming more ubiquitous in almost every sector, it is outpacing the current skills of workers and requiring considerable adaptation on the part of industrial-era employees. According to researchers at Goldman Sachs, it is estimated that these new tools could affect nearly a quarter of current jobs, with the most vulnerable areas being positions in administration, legal services, finance, engineering, and graphic design. The most affected jobs right now are held by women with 80% of their jobs possibly subject to automation, compared to 58% for men.

Computer chips are critically important in each of these developments, and there are particular vulnerabilities owing to the fact that most of the current production takes place in Taiwan and South Korea, places that lie in sensitive geographic areas with lots of geopolitical tensions. If conflict escalates in one or both of these places, it immediately would endanger digital innovation around the world because of the shortages that quickly would emerge. Disruptions of supply chains could be catastrophic for national security and economic prosperity, and pose considerable problems for China, the United States, and European countries.

In response to these risks, the U.S. is seeking to enhance its domestic production capabilities for computer chips. As of 2020, there were around 280,000 jobs in the semiconductor sector with some in the design area but most of the positions in manufacturing. But there are new plants planned in Arizona, Ohio, and New York, and there already are worker shortages reported in those states. Experts estimate that domestically, at least 50,000 engineers and 200,000 to 300,000 additional welders, electricians, mechanics, pipe-fitters, and the like will be needed to push these new facilities across the finish line. No one wants “ghost fabs” that look good on paper, but are not able to produce the number of chips required by today’s digital economy. Annual semiconductor sales around the world total around $580 billion and that number is expected to reach $1 trillion by 2030, making the sector one of the fastest-growing fields.

Difficulties in STEM education and worker training

Currently, the nation is not producing a sufficient number of people skilled in STEM competencies. According to the U.S. Bureau of Labor Statistics, there are 9.9 million STEM workers in America, but that number needs to grow considerably as both the public and private sectors digitize their operations and require highly-trained workers. Native-born American students by and large are not going into these areas in big numbers, and there are limitations on the number of foreign students and immigrants allowed into the country. As an illustration of the problem, 45% of current STEM employees in the U.S. with a doctoral degree are foreign-born, and the backlash against immigration could pose barriers to retaining and growing that talent pool.

Consequently, many individuals have not realized how crucial immigration has been to long-term American innovation. Research by Vivek Wadhwa has shown that half of Silicon Valley companies had an immigrant founder or co-founder. Additional work demonstrates how immigration cutbacks have harmed the ability of high schools to staff their math and science positions. Yet current restrictions on immigration and the ability to get H-1B visas limit the capacity of U.S. firms to hire talented workers to keep the country at the forefront of innovation — in light of domestic skill shortages.

This would be less of a problem if native-born Americans were entering STEM fields, but that generally is not the case. There are barriers facing women in STEM fields that discourage a big part of our potential talent pool from getting advanced degrees in those areas. Only 19.3% of engineering graduates right now are female, and only 17.9% of computer science graduates are women. And some of the same, if not more barriers, impact people of color.

Labor shortages and skills shortfalls pose considerable risks to the future U.S. economy because of their importance for our digital transformation. Additionally, making it difficult for foreign workers to get H-1B visas also impedes U.S. competitiveness by depriving firms of needed workers with the skills required to build semiconductors, staff manufacturing facilities, and provide needed expertise in finance, retail, health, education, and other sectors which are incorporating AI in their operations — all areas of extreme importance to the Biden administration with significant streams of federal investments.

Needed actions

There are several actions that are required to address current challenges and improve worker training. These include effective implementation of the U.S. CHIPS and Science Act, boosting U.S. domestic manufacturing capabilities, better workforce development and STEM education, prioritization of lifelong learning, reasonable immigration policies, and effective management of geopolitical tensions.

Effective implementation of the U.S. CHIPS and Science Act

In recently enacted legislation, the U.S. government is investing $280 billion in efforts to boost computer chip production and innovation. This includes $200 billion in scientific research and development, $52.7 billion in semiconductor manufacturing and development, $24 billion in tax credits, and $3 billion in supply chain management. Owing to the importance of this sector, computer chips have been deemed a national priority and worthy of this extraordinary investment. Among the goals of this legislation is building new chip facilities located in the U.S., improving workforce development, R&D, and providing better management of supply chain issues.

Industry analysts estimate that the $50 billion federal investment in semiconductor manufacturing would add $24.6 billion annually to the American economy and create 185,000 jobs from 2021 to 2026.

Yet achievement of these goals requires effective implementation of this bill. Sophisticated chip fabs require around $20 billion to build. How federal money is spent is crucial to determining whether the bill achieves its objectives. It is important for government and industry to collaborate in order to reach the best results and for the money to avoid siloing problems that focus on narrow goals as opposed to advancing the sector as a whole. For example, the sector needs support in fabrication, assembly, distribution, cybersecurity, workforce development, and supply chain management. Progress in a few of these areas at the expense of the others will not solve the problem.

Making advances in each of these endeavors is crucial to improving the prospects of the sector in general. Industry analysts estimate that the $50 billion federal investment in semiconductor manufacturing would add $24.6 billion annually to the American economy and create 185,000 jobs from 2021 to 2026. Money invested in this area has been projected to have a 6.7 multiplier effect on economic activity, meaning for each worker directly employed by the semiconductor industry, an additional 5.7 jobs are supported in the wider U.S. economy. But currently, domestic workers with the critical skillsets are not fully available to staff the projected job increases, and this could depress economic activity.

To get the greatest bang for the buck, Christopher Thomas of the Brookings Institution proposes a public-private Equity Capital Fund that would have “market-level rates of return and generate significant returns for the government.” The Fund would take a comprehensive approach, leverage private money both from U.S. and non-U.S. firms, and use income gained via that Fund to reinvest in future opportunities that create a self-sustaining source of investment capital. His hope is to spawn multiple investments that would benefit numerous different firms, and thereby maximize the effectiveness of the public money.

Boosting U.S. domestic manufacturing capabilities

With foreign production sites at risk of geopolitical conflict, economic competition, or military action, it is vital to boost U.S. domestic manufacturing capabilities. For years, the U.S. has outsourced production and manufacturing in many sectors to Mexico, China, India, and Vietnam, among elsewhere, and this has left the country dependent on foreign entities for critical goods and services. Now the approach needs to on-shore or near-shore vital sectors to make sure it has what it needs in case of geopolitical conflict or supply chain disruptions. Failure to ensure reasonably located domestic production facilities or those situated among friendly nations could harm the nation’s longer-term security and prosperity.

Better workforce development and STEM education

We need to boost our workforce development and STEM education training in order to stay competitive internationally. Right now, one of the limiting factors in the digital economy is the lack of talent in key science and technology fields. We do not have sufficient native-born workers going into these areas and our national policies are discouraging foreign students and workers from gaining employment in the United States. In earnest, the U.S. could do a poor job in one of those innovation areas and still be okay in the future, but we cannot do poorly in both. It is vital that we encourage native-born Americans to go into STEM fields and that we create pathways for foreign STEM workers to have structured career pathways into these critical fields. Right now, according to a Deloitte study, there are fewer than 100,000 U.S. graduates with electrical engineering and computer science degrees each year, which is below the number that will be required in the coming decade.

If there are few opportunities for women and minorities, we limit the job possibilities for almost two-thirds of the American population, which robs people of economic opportunities but also limits current and future innovation opportunities.

The COVID-19 pandemic did not help with student’s STEM proficiency levels. National Assessment of Educational Progress tests have shown substantial drops in pupil performance, with scores plummeting, especially in math and science. That could affect later student interest in the subject matter and engagement with STEM careers. Some individuals report carryover effects at the college level given a lack of learning critical science and math skills in high school, harming their college performance and career prospects. However, research by Andre Nickow, Philip Oreopoulos, and Vincent Quan shows personal tutoring can help overcome science learning problems and improve student performance across the board.

Part of the overall solution has to be reimagining our tech workforce to become more diverse and inclusive. We need to create opportunities for women and minority students to major in STEM fields and get the training, encouragement, and financial support needed to become successful in these areas. If there are few opportunities for women and minorities, we limit the job possibilities for almost two-thirds of the American population, which robs people of economic opportunities but also limits current and future innovation opportunities.

In a recent policy report, Brookings scholars Annelies Goger and Banu Ozkazanc-Pan suggest a number of steps designed to improve STEM training, such as reducing degree requirements to make the field more accessible to students, allowing credit for prior learning, creating “earn-and-learn” options, and strengthening anti-discrimination monitoring and enforcement. Adopting these actions would encourage greater interest in and provide additional support for those who want to pursue training in science and technology fields.

At the high school level, Massachusetts has garnered interesting gains by making STEM a graduation requirement and implementing proficiency exams. While 12% failed the test on their first try, results improved over time as students became more comfortable with the subject matter and schools adjusted to the proficiency requirement. However, there were differences linked to gender, family income, and English proficiency that shows tests must be designed carefully and not exacerbate existing inequities.

As the digital economy expands, we need people who are skilled in AI, data analytics, machine learning, and cybersecurity, among other areas. There are rising challenges in each area and worker shortages limit our ability to take advantage of these emerging technologies. Remedying those limitations must be a high national priority.

Prioritization of lifelong learning

The American adult education program has been estimated to be around $10 billion in 2016 and is expected to triple in 10 years. The old educational model where people mostly invested in education up through age 25 is giving way to a lifelong learning approach in which people upgrade their jobs skills at ages 30, 40, 50, and 60. With rapid changes in the workforce and the nature of jobs, and a fast pace of technology innovation infusing the workforce, workers need to upgrade their skills throughout their lifetimes.

For this reason, higher education needs to shift its training mission to include adults and workers needing to upgrade their job skills. Brookings Fellow Katharine Meyer shows the continuing benefits of a college education in health and income levels. Yet schools cannot rest on their past laurels but need to diversify their mission and teach adults facing a rapidly changing job landscape. Rather than be content with existing skills, people are going to need to upskill regularly and keep up with emerging technologies, and both community colleges and four-year colleges and universities need to play a major role in training adults. Early research suggests older individuals who take noncredit courses benefit from them and the educational materials improve their job skills and life prospects. With some updating of their curricula and new components on AI, virtual reality, quantum computing, and the metaverse, adult education can become an important contributor to people’s job capabilities.

Reasonable immigration policies

As noted earlier, immigration still is vital to U.S. innovation efforts. A considerable portion of our past innovating advantage stems from foreign students and workers. Slowing the number of foreign students who stay in the U.S. and foreign workers who immigrate here for work will slow future innovation. With few native-born Americans studying STEM fields, at least in the short-term, we will need to rely on those from abroad to maintain global advantage in digital technology. Around 40% of high-skilled semiconductor workers currently were born abroad, making the field also quite dependent on foreign talent.

For foreign students and workers, programs that support their education and help to maintain their working statuses will be keen to these apparent skill and talent shortages among native-born Americans. At least, that should be the case to meet the expectations of the exorbitant federal commitments. With considerable needs for engineers, designers, welders, and pipe-fitters, the country needs to continue what it has been doing for decades, which is filling STEM jobs with workers from abroad who have the skills that native-born students are not developing in sufficient numbers.

Effective management of geopolitical tensions

Tensions between the U.S. and China are at a high level and if they continue to escalate, it will pose major challenges for national security, economic prosperity, and international competitiveness. Leaders in each country mistrust the other and that damages communications, global cooperation, and problem-solving. It leads to disruptions in supply chains, trade disputes, foreign policy disagreements, and possible military or economic confrontations.

It is hard to know how to improve geopolitics given the competing interests of various countries, but it is important to maintain open lines of communications and engage in dialogue that could lead to written agreements or better understandings of the current situation. Letting the situation deteriorate likely will lead to difficult choices for every country involved in global commerce and worse outcomes for leading nations.

Authors

  • Acknowledgements and disclosures

    Thanks to Jack Malamud for providing outstanding research assistance and Katharine Meyer and Nicol Turner Lee for their helpful comments and suggestions on this paper.