History presents us with many doomsday predictions of mass unemployment during periods of technological innovation. The development of robots and generative Artificial Intelligence (AI) seems to be ushering us into another revolution, but the extent to which it is similar to previous revolutions remains unclear. This essay explores three potential economic impacts: unemployment, welfare, and inequality.

Unemployment 

Goldman Sachs et al. (2023) found that “it is estimated that two-thirds of jobs are exposed to generative AI”, around 300 million worldwide. However, AI is also estimated to increase global annual GDP by 7%. We can explore this nuance with the Cobb-Douglas production function.

Production Function Analysis

Modelling the total output of the UK as a function of Capital (K), Labour (L), with total factor productivity (A) gives a Cobb-Douglas production function assuming constant returns to scale:

Deriving marginal productivity of labour (MPL):

Where 0<<1 is capital’s share, the share of national income allocated to capital, (2) shows that income per worker is given by MPL. Crucially, the divide between capital and labour has been historically stable (OECD & Belser, 2015) (Mankiw, 2009).

This function demonstrates important dynamics between labour and capital. Firstly, an increase in K will increase MPL and increase real wages, implying that capital is complementary to labour. Secondly, capital can act as a substitute for labour: assuming Y and A are fixed, increasing K implies L1- decreases. Simultaneously, increasing productivity through technological development (A) will increase GDP (Y). The introduction of AI most definitely increases productivity (A), but the effects on unemployment become murky and circumstantial; it depends if robots and AI are complements or substitutes to labour.

While displacement of workers by robots and AI is inevitable, history suggests that job loss is usually temporary. McKinsey & Company (2017) contends that robots and AI will alter the employment landscape: some jobs will be eliminated, and new ones will be created with different skill requirements. Mudzar & Wai Chew (2022) posit that “high-level technical skills, higher-order cognitive skills, and human or interpersonal skills” will become more valuable, however low or middle-level foundational skills will remain essential. Previous industrial revolutions exhibited similar trends: lower-skilled workers were displaced while higher-skilled workers benefited from new technology.

(Frey and Osborne, 2013)

Nationally, low-skilled workers have to retrain, causing short-term frictional unemployment and long-term structural unemployment. To smooth out this transition, the government can implement training schemes, update educational policy, and provide social care for older workers.

Globally, asynchronous approaches to regulation could lead to failure in moderating transnational activity and protection. In China, AI must be put through a safety assessment before its release, checking whether it undermines state power or encourages discrimination and violence. In contrast, Switzerland only regulates AI in terms of data protection, transparency, and competition, and the UK’s AI regulation is more sector-specific. Shockingly, AI is unregulated in India (Heywood, 2023).

Welfare

AI may not always be damaging; circumstantially, AI cuts down working hours and increases productivity (MPL). For instance, AI saves an attorney's time by taking the burden of legal research, reducing work hours without compromising quality or their ability to remain competitive in the legal field and improve their standard of living. Drastically, AI could significantly boost the quality of service. Notably, Amazon uses AI flywheels that identify consumption patterns, select the most suitable products for recommendation, and adjust pricing using the consumer’s purchasing and viewing history (Bezonomics, 2020).

However, Amazon’s AI flywheels require an extensive database and are initially costly, making them difficult for small businesses to implement. AI’s increasing prevalence may strengthen large monopolies, making it harder for smaller firms to compete. This is a microcosm for global inequality. As developed countries with advanced technology utilise AI to accelerate growth, developing countries with insufficient technology lag further behind.

Inequality

The development of robots and AI widens the gap between those with access to advanced technology and those who do not. Observing previous periods of technological development: monetary indicators suggest that during the early stages of the Industrial Revolution, Britain's wealth and income inequality widened (Lindert, 2000). Allen's (2019) study on the size distribution of inequality in England using the distribution of income in social tables finds an increase in the proportion of income going to the top decile. This is supported by an increase in the Gini coefficient from 0.54 in 1759 to 0.60 in 1798, suggesting high-skilled workers earn higher wages than lower-skilled workers.

Acemoglu and Restrepo (2022) found that “between 50% and 70% of changes in the U.S. wage structure over the last four decades are accounted for by relative wage declines of worker groups specialised in routine tasks in industries experiencing rapid automation”, indicating that the rise of AI may have similar effects on income and wealth distribution. Deloitte (2017) claimed that 39% of all legal jobs are at risk of automation, which would have two primary impacts: lower demand for legal skills and an increase in profit due to the lower wage costs brought about by technological improvement. This captures the essence of introducing robots and AI: lower-skilled paralegals risk being substituted, while the aforementioned higher-skilled attorneys become more productive.

Closing remarks

From historical observations, the short-term economic impacts of robots and AI will be harsh, but markets will eventually adapt to new technology. Labour impacts depend on the sector and skill: wages increase if AI and labour are complementary and decrease if they are substitutes. Thus, demand for highly skilled labour increases. Equally, welfare increases as robots and AI increase total factor productivity, increasing the quantity and quality of goods and services. However, inequality increases due to productivity differences between individuals and countries with and without access to such transformative technology. As with any change, the future is uncertain; governments must devise strategies to prepare for the impacts of robots and AI, minimise economic damage, and welcome the revolution.

Bibliography

Acemoglu, D. & Restrepo, P. (2022) Tasks, Automation, and the Rise in U.S. Wage Inequality. The Econometric Society. 90 (5). 

Allen, R.C. (2019) Class structure and inequality during the Industrial Revolution: lessons from England’s social tables, 1688–1867. The Economic History Review. 72 (1), 88–125. doi:10.1111/ehr.12661. 

Dumaine, B. (2020) Bezonomics: How Amazon Is Changing Our Lives, and What the World’s Companies Are Learning from It. Simon and Schuster. 

Financial Times & McDermott, J. (2014) Is your job safe in the second machine age? 2014. Financial Times. https://www.ft.com/content/c8901cc7-d879-3fb7-89ea-16aab9bec3e7 [Accessed: 13 July 2023]. 

FRED (2021) Share of Labour Compensation in GDP at Current National Prices for the United States. 21 January 2021. St. Louis Fed. https://fred.stlouisfed.org/series/LABSHPUSA156NRUG# [Accessed: 2 July 2023].

Frey, C.B. & Osborne, M.A. (2013) The Future of Employment: How Susceptible Are Jobs to Computerisation? 1 January 2013. Elsevier. https://www.researchgate.net/publication/271523899_The_Future_of_Employment_How_Susceptible_Are_Jobs_to_Computerisation.  

Goldman Sachs, Briggs, J., Kordani, D., Hatzius, J. & Pierdomenico, G. (2023) Global Economics Analyst The Potentially Large Effects of Artificial Intelligence on Economic Growth (BriggsKodnani). https://www.key4biz.it/wp-content/uploads/2023/03/Global-Economics-Analyst_-The-Potentially-Large-Effects-of-Artificial-Intelligence-on-Economic-Growth-Briggs_Kodnani.pdf.  [Accessed: 2 July 2023].

Haggerty, J. (2017) Disruptive technology in the legal profession. 15 June 2017. Deloitte United Kingdom. https://www2.deloitte.com/uk/en/pages/financial-advisory/articles/the-case-for-disruptive-technology-in-the-legal-profession.html [Accessed: 11 July 2023]. 

Heywood, D. (2023) The UK’s approach to regulating AI. 2023. TaylorWessing. https://www.taylorwessing.com/en/interface/2023/ai---are-we-getting-the-balance-between-regulation-and-innovation-right/the-uks-approach-to-regulating-ai [Accessed: 11 July 2023]. 

Lindert, P. (2000) When did inequality rise in Britain and America? Journal of Income Distribution. 9 (1), 11–25. doi:10.1016/s0926-6437(99)00012-8. 

Mankiw, N.G. (2009) Macroeconomics. Worth Publishers. https://jollygreengeneral.typepad.com/files/n.-gregory-mankiw-macroeconomics-7th-edition-2009.pdf

Manyika, J., Lund, S., Chui, M., Bughin, J., Woetzel, J., Batra, P., Ko, R. & Sanghvi, S. (2017) James Manyika. 28 November 2017. McKinsey & Company. https://www.mckinsey.com/featured-insights/future-of-work/jobs-lost-jobs-gained-what-the-future-of-work-will-mean-for-jobs-skills-and-wages [Accessed: 11 July 2023]. 

Mudzar, N.M.B.M. & Wai Chew, K. (2022) Change in Labour Force Skillset for the Fourth Industrial Revolution: A Literature Review. 2022. IJTech - International Journal of Technology. https://ijtech.eng.ui.ac.id/article/view/5875 [Accessed: 9 July 2023]. 

OECD & Belser, P. (2015) The Labour Share in G20 Economies. https://www.oecd.org/g20/topics/employment-and-social-policy/The-Labour-Share-in-G20-Economies.pdf.   

Vipond, H. (2022) Technological Unemployment in Victorian Britain. 1 June 2022. Economic History. https://blogs.lse.ac.uk/economichistory/2022/06/01/technological-unemployment-in-victorian-britain/ [Accessed: 9 July 2023].