FOR IMMEDIATE RELEASE

CONTACTS

Dr. Carl Van Horn, Distinguished Professor and Director, vanhorn@rutgers.edu / (848) 932-1153

Robb C. Sewell, Assistant Director of Public Relations and Communications, rcsewell@ejb.rutgers.edu / (609) 217-8697 (mobile)

The topline with methodological statement is available at either of the following links:

https://go.rutgers.edu/h281hqnh

https://heldrich.rutgers.edu/sites/default/files/2024-02/Work_Trends_February_2024.pdf

Data visualizations are available here:

https://go.rutgers.edu/yi77szhg

https://heldrich.rutgers.edu/sites/default/files/2024-02/Work%20Trends_February_%202024_AI_Infographics_0.pdf

U.S. Workers Assess the Impacts of Artificial Intelligence on Jobs

Government Regulation of AI Supported by Democrats, Republicans, and Independents

New Brunswick, New Jersey (February 7, 2024) - In a national survey of U.S. workers conducted by the Heldrich Center for Workforce Development at Rutgers, The State University of New Jersey, 8 in 10 (82%) workers say it is very or somewhat important that the government in Washington, D.C. ensures that artificial intelligence (AI) technologies do not cause U.S. workers to lose their jobs.[1] Nearly half (46%) say it is very important.

• Republicans, Democrats, and Independents hold similar views about the role of government when it comes to AI and jobs — at least 4 in 10 say it is very important (45%, 51%, and 43%, respectively).[2]
• The nation’s lowest-income earners — those making less than $50,000 annually (54%) — and workers without a four-year degree (53%) — are more likely than those earning more than $100,000 annually (35%) and workers with a four-year degree (36%) to say they think the government’s actions that would limit the impacts of AI on jobs are very important.[3]
• Seven in 10 U.S. workers say they are very or somewhat concerned about employers using AI in human resources decision-making (71%), 5 in 10 are concerned about needing more technological skills (50%), and 3 in 10 worry about their jobs being eliminated (30%).

The Heldrich Center’s probability-based survey of 1,038 U.S. adults — including 737 workers — found widespread uncertainty about the potential impacts of AI on jobs and the labor market.[4] In a year that may be pivotal for the regulation of AI, these findings demonstrate important public support for the development of AI governance and guidance for how employers deploy AI technologies in the workplace.[5]

President Joe Biden’s executive order signed in late 2023 directs federal agencies to develop guidelines for the responsible use of AI by private companies, academia, and governmental agencies. It also contains a “commitment to supporting U.S. workers,” including assessing job displacement risks due to the application of AI in the workplace.

Concerns about the impact of new technologies on jobs are not new for workers. The Heldrich Center’s 2018 Work Trends study found that workers living in lower-income households are more likely to express worry about the impacts of technology on work and the economy, compared with workers living in higher-income households. The high level of concern at this early stage in the application of generative AI is noteworthy.

“As with other major technological changes, generative AI will create opportunities for some and heartbreak for others. Workers — especially those with the least formal education — want safeguards that protect them from disruption and unemployment,” said Carl Van Horn, Ph.D., Distinguished Professor of Public Policy and Director of the Heldrich Center.

While the survey showed that 71% of the U.S. labor force is concerned about the use of AI in employers’ decision-making about hiring and/or promotions, this concern ranks lower than concerns about the cost of living (94%) or fears of an economic recession in 2024 (83%).

Workers living in lower-income households are twice as likely to say they are very concerned about AI and economic issues compared with workers in higher-income households (39% vs. 17%). More females (78%) than males (64%) express concern about AI and jobs. Four-year degree holders are less likely to say they are concerned about technology’s impact on jobs in general, compared with workers without a four-year degree (53% vs. 70%). When asking about AI, however, there are no differences by education (67% college graduates, 73% less than college).[6]

Most U.S. workers indicate that they heard a little about the impact of AI on workers and jobs in the United States in 2023 (64%). One in 10 have not heard, read, or know a lot about how AI will affect jobs and workers. Seven in 10 U.S. workers say they know a little about how AI may or may not change the nature of jobs in the U.S. labor market (68%).

Most U.S. workers say they are not concerned that AI will cause their own jobs to be eliminated but worry about the impact of AI on employers’ hiring and promotion decision-making. Because of the use of AI by employers now or in the future, 71% of the U.S. labor force say they are very (26%) or somewhat (45%) concerned about the possibility that employers use AI when making decisions about hiring and/or promotions. Workers of color (34%) are also more likely to say they are very concerned about the use of AI in employer decision-making than White workers (19%).[7] There are no meaningful differences by household income, gender, or educational attainment when asking about the use of AI in decision-making about hiring and promotions.[8]

Thirty percent (30%) of U.S. workers[9] say they are very (13%) or somewhat (17%) concerned that their own jobs will be eliminated. Job displacement is a concern for:

• Workers living in lower-income households: 45% of workers earning less than $50,000 annually are concerned about their own job being eliminated (23% are very concerned), compared with 18% of workers earning more than $100,000 annually (4% are very concerned).
• Workers of color: one in four (24%) workers of color say they are very concerned about their jobs being eliminated due to AI, compared with 1 in 10 White, non-Hispanic workers (6%).
• Workers without a college degree: 36% of workers who have completed some college or less say they are very concerned, compared with 20% of workers with a four-year degree.

Fewer than 1 in 10 (9%) U.S. workers say they agree a lot with the statement that “artificial intelligence is good for the economy.” Sixty-three percent (63%) of workers with a college degree say AI is good for the economy, while 43% of workers without a degree say the same. Workers of color are surprisingly also more likely to agree compared with White workers (59% vs. 45%). Six in 10 males and 4 in 10 females agree (59% vs. 44%).

Conversely, an August 2018 Work Trends study[10] about workers’ experiences with technology found that 43% of the U.S. labor force said they agreed a lot with the statement “new technologies are good for the economy.”[11]

When asked about the impact of any “new technologies”[12] vs. “AI” on job displacement in 2023, there are no differences in the views expressed by the Heldrich Center’s survey respondents. Sixty-seven percent (67%) of workers say AI will eliminate more jobs that it will create, while 63% of workers say new technologies eliminate more jobs than they create. There are also no meaningful differences when asking about new technologies between 2018 and 2023.

Workers without a four-year degree are more likely than degree holders to say that new technologies eliminate more jobs than they create (71% vs. 52%) and AI will eliminate rather than create jobs in the future (75% vs. 55%). Thirty-one percent (31%) of workers living in lower-income households say they agree a lot that AI will displace workers, compared with 13% of workers living in households earning more than $100,000 annually.

While most U.S. workers say they agree that AI will eliminate, rather than create jobs, they are divided when asked about who may be more likely to lose their jobs. Forty-six percent (46%) of U.S. workers agree a lot or a little with the statement “higher-income workers are more likely to lose their jobs because of artificial intelligence, compared to lower-income workers.” Fifty-three percent (53%) say they disagree a little or a lot. Higher-income workers categorically disagree that jobs like their own are more likely to be eliminated, compared with lower-income workers. Five in 10 workers earning less than $100,000 annually agree with the statement (54% less than $50,000 annually and 49% $50,000 to $100,000 annually), compared with 3 in 10 workers living in households earning more than $100,000 annually (33%). Half of workers who have completed some college (53%) agree a lot or a little that higher-income workers will face job displacement. Just 35% of college completers say the same.

Most workers (66%) say they agree with the statement that, in general, “they will need more technological skills to achieve their career goals.”[13] The Heldrich Center has observed no change on this measure since 2018, providing evidence that workers were no more or no less concerned about upskilling because of technology in general in 2023 than they were in 2018. However, because of the application of AI at work, about half (50%) of the U.S. labor force say they are very (14%) or somewhat (36%) concerned that they will need more technological skills to achieve their career goals. These opinions differ by household income and education: 58% of workers living in households earning less than $50,000 annually and 55% of workers without a four-year degree say they are very or somewhat concerned they will need more skills, while 41% of households earning more than $100,000 annually and 42% of degree holders say the same. Workers of color are nearly three times more likely to say they are very concerned they need more technological skills to achieve their career goals compared with White workers, and two times more likely to agree a lot that they need more technological skills to achieve their career goals (22% vs. 8% and 23% vs. 11%). 

Notes

[1] U.S. workers are employed full time, employed part time, unemployed and looking for work, retired but now working, and on temporary furlough from a job (N=737).

[2] Party identification is tabulated without leaners. Respondents were asked if they are Republican, Democrat, or Independent. Respondents who indicate they are “something else” were tabulated with Independents. 

[3] See topline for data disaggregated by demographic subgroups.

[4] The following definition of AI was provided to respondents when it was first mentioned in the survey: Artificial intelligence is the use of machines or technology to complete tasks by simulating human intelligence.

[5] A 2023 Washington Post-Ipsos poll reported that most workers don’t think their jobs will be displaced before 2040, and the Pew Research Center found that while workers are concerned about some uses of AI by employers, they also approve of others.

[6] In this study, when disaggregating data by respondents’ educational level, those having attended at least some college and those with a high school degree or less tended to have similar opinions overall. There are a few instances in which those with some college education have opinions similar to four-year college degree holders. When disaggregating data by respondents’ educational level using three groups (high school or less, attended some college or completed an associate degree, and four-year degree holders and more), fewer four-year degree holders (7%) and individuals with some college (12%) say they know nothing at all about how AI may or may not change the nature of jobs in the United States compared with those with a high school degree or less (22%). In addition, four-year degree holders (35%) and those with some college (43%) are less likely than those who graduated high school or less (61%) to agree that higher-income workers are more likely to lose their jobs because of AI, compared with lower-income workers. Finally, U.S. adults with a high school degree or less (32%) are more likely to be very concerned about the general impact of technology on jobs than adults who attended some college (24%) and college graduates (13%).

[7] Workers of color are represented as BIPOC in the topline (Black, Indigenous, and People of Color).

[8] “Meaningful” means outside of the margins of error for each study. See margins of error reported in the methodology.

[9] Employed U.S. residents age 18+ (N=677).

[10] Includes employed full time, part time, unemployed and looking for work, and military. The 2018 survey did not explicitly ask about retired but working.

[11] “New technologies” was the term used in 2018. For data based on the labor force, visit the Roper Center’s iPoll archived data: https://ropercenter.cornell.edu/ipoll/study/31116367.

[12] In 2018, respondents were provided with the following prompt, prior to the survey questions: Over the past 20 years, developments in technology, such as computers, have changed many aspects of the U.S. economy, including the availability of goods and services to consumers and the nature of jobs in the workplace. Advances in these new technologies, including widespread use of the Internet, automation of job tasks and processes, and the use of robots, are changing how people work. In 2023, respondents were provided with the following prompt, following the Q5 battery but before all other questions: Over the past 20 years, developments in technology, such as computers and the Internet, have changed many aspects of the U.S. economy, including the nature of jobs and the availability of goods and services to consumers. Advances in these new technologies, including the automation of job tasks, are changing how people work.

[13] See topline for question wording. The survey asked respondents this question twice. First, in the same battery as the question asking about “new technologies.” The second time was prefaced with the prompt, “because of the use of artificial intelligence by employers now or in the future.”