The Principle of Comparative Statics in Labor Markets: Analysing the Impact of AI on Employment

As we saw in our analysis of post-COVID shifts in the housing markets, the Principle of Comparative Statics examines the effects on a market due to a change in some exogenous factor. Artificial Intelligence (AI) is an external factor currently having a major impact on the labour market. Some people worry that it will ultimately shake the labour markets more than anything to date, in particular swallowing up many of the knowledge jobs created over the last 20 years.

The Impact of AI on the Labour Market

The Tony Blair Institute for Global Change released a report on The Impact of AI on the Labour Market in 2024. They concluded that “full and effective adoption of AI by UK firms could save almost a quarter of private-sector workforce time – equivalent to the annual output of 6 million workers.” If you scale this across the entire globe (or at least the parts of it with access to modern technology) you can see a massive shock to the world’s labour force.

The report writers observe that the potential for AI to boost productivity will differ at the individual worker level, however, with most savings coming from AI-software performing cognitive tasks, rather than physical tasks. Workers performing routine cognitive tasks are the most exposed.

They recognize that although some workers will lose their jobs to AI, the technology will also create jobs by boosting economic growth and speeding up the development of new goods and services. They see AI continuing the push and pull between labour substitution and demand creation.

In all the reportwriters’ scenarios, unemployment will initially rise as firms reduce their workforce. This should happen gradually, alongside the uptake of AI over the wider economy. In the case of the UK, they see AI supplanting 60,000 to 275,000 jobs a year. Many of these workers have the potential to rejoin the workforce, however, in new industries created.

In addition, the report views that AI has the potential to improve the supply of labour in three main ways:

1. AI in the classroom has the potential to raise educational achievement. The report writers estimate that AI can improve student attainment by 6%
2. AI can potentially improve the health of the workforce, with fewer lost workdays
3. AI can support better job-matching, leading to more effective utilization of labour

OECD Findings About the Impact of AI on the Labour Market

The OECD released a working paper on The Impact of Artificial Intelligence on the Labour Market. The writers believed that the impact of AI is ambiguous, depending strongly on factors like the type of AI being developed and deployed, how it is developed and deployed, and market conditions and policy. However, empirical evidence of AI implementation to date suggests there will not be major job losses and declining wages. Much of the impact of AI on jobs comes through the reorganisation of tasks within an occupation.

The report writers observe that workers need to re-skill or up-skill to adapt to the reorganisation of tasks and the emergence of new tasks.

The Impact on Individual Workers Can Be Very High

Researchers at the University of Otago (on behalf of the New Zealand Law Foundation) tried to predict the impact of AI on jobs in New Zealand. They saw little point in trying to predict the number and types of workers whose jobs will be impacted by AI. There are too many decisions yet to be made. However, although widespread unemployment is unlikely, the cost of unemployment for individuals is so high that “most will choose even low-value, precarious work over no work at all.”

Shift of Labour Demand Across Industries Due to AI Adoption

Artificial Intelligence's Creation and Displacement of Labour Demand by Taelim Choi and Nancey Green Leigh explored the dynamics of labour demand creation and displacement from adopting AI in US metropolitan statistical areas. Their analysis shows that AI technologies are increasingly penetrating major industries and disproportionally generating new labour demand for AI-skilled workers in the regions where AI innovation occurs. They found demand for non-AI labour declines slightly in sectors and areas with higher AI skill adoption rates.

The Impact of ChatGPT on the Labour Market

A common thread through all the above studies is that it is too early yet to have a definitive understanding of how AI will affect employment. Ali Zarifhonarvar, a Visiting Research Scholar at Indiana University, has looked at an existing AI technology in the Economics of ChatGPT published in the Journal of Electronic Business & Digital Economics. He found that 32.8% of occupations could be fully impacted by ChatGPT, while 36.5% might experience a partial impact and 30.7% are likely to remain unaffected. He does qualify that these figures are predictions – even with ChatGPT, data showing actual effects doesn’t exist yet.

Like most of the other researchers, Zarifhonarvar found that the short-term impact of ChatGPT and other generative AI services on the job market is unpredictable and might go either way. There will be a temporary mismatch between the skills employees possess and the capabilities employers need.

He viewed the short-run impact of AI on labour demand and supply is as follows, assuming a relatively inelastic labour supply curve.

The long-run effect of ChatGPT and other generative AI is less predictable, but Zarifhonarvar sees two possible scenarios:

1. The increase in AI services could result in an increase in the total number of job vacancies as well as an increase in the wage for those positions, due to increased productivity. This would lead to economic growth and increased labour demand. In this case, employment and wages would rise in the long term. Alternatively,
2. The increased level of automation that generative AI brings about will reduce the demand for human workers. In this case, the decreased demand for labour will reduce employment and wages (for certain workers at least).

The jobs suffering the greatest impact from AI are coders, software developers, computer programmers, and data scientists, as well as media jobs, including those in advertising, technical writing, journalism, and content creation. Other areas affected include legal industry jobs such as paralegals and legal assistants, customer service representatives, teachers, those working in certain finance-related industries, and market research analysts.

There is likely to be polarization in labour markets, with AI having minimal effect on both workers with high job skills and those with low job skills. However, those with medium skill levels (as listed above) are most likely to feel the impact of AI.

How Governments Can Mitigate the Effects of AI on the Workforce

In the current political environment, it is debatable whether governments will be willing to mitigate the effects of AI on the workforce, particularly if it involves spending money. However, there are some courses of action that governments could consider.  

They could provide incentives to help AI-replaced workers to reskill, providing workers with the skill sets currently in demand. For example, in South Carolina, businesses that sponsor a registered apprenticeship program can receive a $1,000 direct tax credit for each registered apprentice employed for at least seven months during each year of apprenticeship.

Governments could offer some form of AI Adjustment Assistance, offering partial unemployment compensation to employees facing reduced work hours. Again, this would probably dissipate over time, allowing workers to upskill themselves.