What is Technological Unemployment?

What Is Technological Unemployment?

Technological unemployment is the displacement of workers by machines. When a new technology performs a task that humans used to do — and there is no equivalent task waiting for those workers — the result is unemployment that is not caused by a business cycle or a downturn but by the technology itself. The term was coined by the economist John Maynard Keynes in 1930, who described it as "unemployment due to our discovery of means of economising the use of labour outrunning the pace at which we can find new uses for labour."

The concept is easy to misunderstand. It does not mean that every worker replaced by a machine loses their livelihood forever; it describes the mechanism of displacement, not its final outcome. A bank teller replaced by an ATM may become a customer service representative, or may not. The question that has haunted economics for two centuries is whether the mechanism of displacement ultimately leaves more people worse off than better off — or whether the system, left to itself, absorbs the displaced into new work.

What makes the topic urgent now is the nature of the current technology. Previous waves of automation primarily replaced physical labor and routine cognitive work. AI, especially generative AI, performs tasks at the heart of cognitive and creative work — writing, analysis, translation, design, code. The question is whether this time the "new uses for labour" will appear fast enough, and whether the workers displaced will have access to them.

Historical Background

The fear of machine-caused unemployment is as old as the Industrial Revolution. In 1811, the Luddites smashed textile machinery they believed was destroying their livelihoods. For most of the nineteenth century, many economists accepted the possibility of long-run technological unemployment, even as the industrial economies expanded. The concept moved into the mainstream with Keynes's 1930 essay, which predicted that within a century, technological progress would solve the "economic problem" of scarcity and leave humanity with the problem of leisure.

The twentieth century tested that prediction and seemed to refute the pessimists. Each wave — the assembly line, electrification, computing — eliminated whole categories of jobs, yet employment rates rose over the long run because new industries emerged, productivity gains were spent on new goods and services, and demand for labor grew in sectors like healthcare, education, and services. The economist's term for this resilience is the "compensation theory" of technological change: displacement today, compensation tomorrow.

The modern debate reopened in the 2010s. Frey and Osborne's 2013 study estimated that nearly half of US jobs were at high risk of computerisation, and the OECD produced more moderate but still significant estimates. The 2020s wave of generative AI prompted economists like Acemoglu and Restrepo to investigate empirically how automation actually affects local labor markets, finding that industrial robots reduce employment and wages in the communities where they are adopted — while also documenting that the effects are far from the apocalypse scenario.

Key Concepts

Displacement is the direct effect: the task is gone, and with it the job that consisted of that task. Displacement is easy to measure in specific industries and hard to measure in the aggregate, because workers move, occupations change, and new tasks appear.

Compensation is the countervailing mechanism. Displaced workers are reabsorbed through several channels: new occupations created by the technology itself, lower prices that raise real incomes and create demand, and shifts of labor into non-automatable services. Whether compensation keeps pace depends on the elasticity of demand for new goods and services — and that is an empirical question, not a certainty.

Skill-biased change is the distributional twist. Automation raises demand for high-skill workers who complement the technology and lowers demand for workers whose tasks are substituted. The result is not uniform unemployment but a hollowing out of middle-skill jobs, a polarization of the labor market, and wage stagnation for workers whose skills compete with machines.

The AI difference is the new question. Earlier technologies replaced specific tasks within well-defined domains, leaving vast areas of work untouched. AI is a general-purpose technology that can be applied to almost any cognitive task, and it improves over time. Whether the compensation mechanisms of the past — new industries, new tasks, new demand — can absorb a technology of this breadth is genuinely uncertain, and serious economists disagree.

The institutional variable is the wildcard. Technological unemployment becomes mass suffering only when displaced workers have no pathway to new livelihoods. Education, retraining, social protection, labor bargaining, and macroeconomic management all determine whether technological change is a rising tide or a rip current. The same technology can produce very different outcomes in different institutional settings.

Contemporary Relevance

The AI wave is already producing real displacement in specific sectors: call centers, translation, routine coding, data entry, and entry-level content production. At the same time, employment in most advanced economies remains high, and studies of AI adoption in the workplace often find augmentation — workers using AI to do more — rather than pure substitution. The pattern is patchwork: displacement concentrated in some roles, transformation everywhere, and new categories of work emerging around the technology itself.

Policy responses are taking shape. Retraining programs are being expanded, though their record is mixed. Some governments are experimenting with wage insurance, portable benefits, and universal basic income. The European Union is regulating AI in the workplace, including requirements for transparency and human oversight in decisions affecting workers. The debates reveal a shared recognition that the market alone may not absorb the displaced quickly enough.

For individuals, the practical response to technological unemployment is the same whether one is optimistic or pessimistic: diversify skills, understand the technologies that are reshaping one's field, and build capabilities that machines do not yet have — judgment, trust, creativity, and the ability to work with AI rather than against it. For societies, the answer is institutional: the historical lesson is not that technology always provides, but that societies that invested in education and social protection weathered technological storms far better than those that left workers to fend for themselves.

Sources

• Acemoglu, Daron, and Pascual Restrepo. Robots and Jobs: Evidence from US Labor Markets. Journal of Political Economy, 2020. https://doi.org/10.1086/705716
• Stanford Encyclopedia of Philosophy. Philosophy of Economics. https://plato.stanford.edu/entries/economics/

Related Topics

• Will AI Replace Jobs?
• What Is the Future of Work?
• What Is Automation?
• What Is the Second Machine Age?
• Automation Topic
• Philosophy of Economics
• Future of Work Topic