SOFIE PEDERSEN – 2024
Like a science fiction movie plot, artificial intelligence (AI) is living up to its name: artificial yet capable of human intelligence, now ready to usurp human jobs. The International Business Machines Corporation’s (IBM) fourth-quarter earnings report in March announced that they plan to lay off 3,900 people in order to focus on AI. IBM is not alone; during January and February 2024, 15,225 cuts were explained as a “technological update,” a discrete term to disguise AI’s causation. Before 2024, only 264 cuts were correlated to this reason, marking a sharp incline in AI-related layoffs. AI will undoubtedly reshape the labor market, and although risks of reduced worker welfare are high, it also has the potential to transform the size and
production of the economy.
AI’s unemployment threat poses issues beyond displaced workers. A decreased labor share in the economy exacerbates income inequality, as capital generally rests in the hands of the wealthy. In general, unemployment harms the economy by reducing the purchasing power of those affected, curtailing consumer spending that represents 70% of the United States (US) gross domestic product (GDP). From a welfare perspective, unemployed workers have worse mental and physical health, in part due to how 75% of professionals gain healthcare through their work.
In April, Goldman Sachs determined that two-thirds of US jobs are vulnerable to AI, with up to 50% at risk of replacement. White-collar jobs with primarily prediction and data analysis functions have the highest exposure. Predictions are necessary for any decision-making, meaning areas of impact can range from a judge’s outcome in a court case to a manager’s choice to hire
workers.
While those numbers may sound intimidating, AI also has the potential to counterbalance its downsides. A study by Goldman Sachs found AI integration could lead to a 7% increase in global GDP and a 1.5% boost in productivity during the next ten years. If AI strengthens productivity, production costs will be reduced, thus increasing labor demand and wages. The World Economic Forum predicted in 2020 that AI would globally displace 85 million jobs by 2025 but establish 97 million new roles. This course of events would not be exceptionally unusual; 60% of current US jobs did not exist in 1940. MIT Economics Professor Daron Acemoglu and Boston University’s Pascual Restrepo compared AI to the automation that occurred post-World War II. Its dual displacement and introduction of jobs resulted in a net positive, with higher employment and wages following the war.
Of course, none of these benefits are guarantees. Large technological companies, like IBM, steer the future of AI — and currently have business models prioritizing AI as a substitution for labor over a productivity enhancer. The government should step in and design tax systems that incentivize human-centered AI. Caused by reforms between 2000 and 2018, the opposite is currently the case: by taxing labor heavier than capital investment, the US encourages automation. In research by Acemoglu, International Monetary Fund’s Andrea Manera, and Restrepo, they determined the US effective tax rate on labor to be 25.5-33.5% while capital taxes are only 5% to 10%. More ideal effective tax and capital tax rates would be 27% and 8%, respectively; this would raise employment by 4% and labor share income by 1%. As another option, the researchers suggest a 10.15% automation tax on tasks where humans have a comparative advantage. In other words, only roles where capital replacement lowers unit costs by more than 10.15% should be substituted for AI.
However, as workers will inevitably continue to be displaced by AI, tax systems are not sufficient to maintain welfare. Subsidizing retraining and temporary employment programs is a crucial. Upskilling programs should focus on retraining workers in a new discipline of their industry, thus designing a future that allows professionals to continue in their field alongside AI. Harvard’s Lawrence F. Katz, Brown’s Jonathon Roth, and research company MDRC’s Richard Hendra and Kelsey Schaberg examined the results of sector-based retraining programs in areas like healthcare, IT, and manufacturing. They found that the initiatives resulted in a wage increase of between 14% to 38%. While there is potential for both positive and negative, AI integration in the economy is still largely uncertain. That said, AI will have a place in businesses, so workers can choose to stand still or stand by it — to let it pass them by or to adapt and grow with AI.
As AI is progressively adopted and its capability increases, productivity grows alongside labor displacement, suggesting that full AI integration cannot be without significant unemployment (Goldman Sachs, 2024).
Bibliography
Acemoğlu, Daron. Harms of AI. The Oxford Handbook of AI Governance. Oxford University
Press, 2023. Accessed March 15, 2024.https://doi.org/10.1093/oxfordhb/9780197579329.013.65.
Agrawal, Ajay, Joshua S. Gans, and Avi Goldfarb. “Artificial Intelligence: The Ambiguous Labor Market Impact of Automating Prediction.” Journal of Economic Perspectives 33, no. 2 (2019): 31-50. Accessed March 15, 2024. https://pubs.aeaweb.org/doi/pdfplus/10.1257/jep.33.2.31.
“AI’s impact on labor productivity depends on its capability and adoption timeline.” Infographic.
Goldman Sachs. April 5, 2023. Accessed March 15, 2024. https://www.goldmansachs.com/intelligence/pages/generative-ai-could-raise-global-gdp-by-7-percent.html.
Bohn, Katie. “Unemployment associated with worse mental and physical health later in life.” Penn State. Last modified June 22, 2022. Accessed March 15, 2024. https://www.psu.edu/news/research/story/unemployment-associated-worse-mental-and-p
hysical-health-later-life/.
Briggs, Joseph, and Devesh Kodnani. “Generative AI could raise global GDP by 7%.” Goldman Sachs. Last modified April 5, 2023. Accessed March 15, 2024. https://www.goldmansachs.com/intelligence/pages/generative-ai-could-raise-global-gdp-by-7-percent.html.
Dao, Mai Chi, Mitali Das, Zsoka Koczan, and Weicheng Lian. “Drivers of Declining Labor Share
of Income.” International Monetary Fund. Last modified April 12, 2017. Accessed March
15, 2024. https://www.imf.org/en/Blogs/Articles/2017/04/12/drivers-of-declining-labor-share-of-inc
ome#:~:text=The%20result%20is%20that%20a,likely%20to%20raise%20income%20ine
quality.
Gil, Bruce. “IBM is cutting jobs as tech’s shift to AI continues.” Quartz. Last modified March 13, 2024. Accessed March 15, 2024. https://qz.com/ibm-tech-layoffs-ai-1851330179.
“Job Cuts Jump in February 2024; YTD Cuts Down 8% Over Last Year.” Challenger, Gray, & Christmas, Inc. Last modified March 7, 2024. Accessed March 15, 2024. https://www.challengergray.com/blog/job-cuts-jump-in-february-2024-ytd-cuts-down-8-o
ver-last-year/.
Stropoli, Rebecca. “A.I. Is Going to Disrupt the Labor Market. It Doesn’t Have to Destroy It.” Chicago Booth Review, November 14, 2023. Accessed March 15, 2024. https://www.chicagobooth.edu/review/ai-is-going-disrupt-labor-market-it-doesnt-have-de
stroy-it.
