Automation and Employment
– A Historical Analysis
von Arnold Kinzel
Consultant | Advisory
19/03/2019 | Frankfurt am Main
Automation is defined as the substitution of human work by machines.
Following this definition, automation is not restricted to manufacturing but may apply to bureaucratic processes, hence including environmental factors, as well.
Back in the day, when the first steam engines were developed, people became worried about their employment status. It was assumed that unemployment would increase vis-à-vis mechanization thus potentially causing a demand shock and economic downturn. Instead, the western world has experienced constant GDP growth and rising prosperity. With the gold standard as the dominant monetary system, real wages have grown substantially from the 1820s until the early 1900s.
Current trends in automation and digitization once again fuel movements against this technological revolution. Academics and politicians discuss possible solutions such as universal basic income or an additional tax on companies with a certain threshold of implemented automation. But is such fear justified or is the market capable of finding an equilibrium where employment remains low and economic growth constant?
A brief history of automation
During the industrial revolution, agriculture made up 41 percent of employment (Autor, 2014). With the development of engines and thereby vehicles and machines supporting harvesting, the industry became more efficient and the required workforce for producing a given amount of crop declined significantly to two percent.
And yet, during this period, employment remained stable and economic output increased constantly. The economic intuition for this phenomenon is explained by marginal substitution rates. During the 1900s, trade in the western world has grown and comparative advantages were being exploited accordingly. Marginal productivities of labor and capital differed greatly across different countries. Nations with sparse populations have had a high marginal rate of productivity for every new worker in the labor force and countries with large populations a low marginal rate. The latter case concerned primarily developed countries where land was expensive and thus labor-intensive industries faced difficulties to develop (Graff, Kenwood, & Lougheed, 2014).
After the introduction of machines into agriculture, the industry became less labor-intensive and products thus cheaper. This led to cost reductions by exploiting economies of scale and created new demand in other sectors where production was potentially more
labor-intensive. Hence, the market found a new equilibrium where economic growth remained at a natural level while everyone was better-off by cost reductions.
On the future of employment
The relation described above triggered a cycle of innovation which is known today as the industrial revolution. Agriculture became more automized while more labor was channeled to the producing industry. Is increasing automation nowadays triggering another revolution or are negative effects dominating?
A similar division can be observed today between manufacturing and the service sector. While manufacturing processes become increasingly more automated, the service sector has grown its share of total economic output. While it is predicted that 47 percent of all jobs in the U.S. are at risk (Frey & Osborne, 2017), a similar situation was observed in agriculture in the 19thcentury. Free global trade has guaranteed the optimal exploitation of comparative advantages which furthermore has helped to channel jobs where demand was created. Free labor mobility and trade are two important factors for such a structural change, both which have never been as extraordinarily present as today, especially in the service industry.
Against the current political perception, automation bears the opportunity to create more jobs in developed economies. In fact, the World Economic Forum has forecasted that automation will create up to 21 million jobs within the next years accounting for losses due to automation (Cognizant, 2019). Although less labor is required for production, more skilled labor will be demanded for the maintenance of automated systems. In addition, products where time-to-market delivery is important might move back from the periphery to markets with a high capitalization and wealthier customers. Labor-intensive production, where the costs of automation do not justify the savings, will still be located in emerging countries with lower wages.
A factor to consider in this analysis are sunk costs. Companies have already moved a large share of their production to peripheral economies during times when the potential of automation was not clear yet. Thus, it is indeed rather ambiguous where skilled labor will move in the future, however, not that more creative and skilled labor will be required to face future innovation challenges.
And yet, it is known that economic and political equilibrium do not always coexist simultaneously. Following the assumption that history repeats itself in some way, one may draw comparisons to the past. It has been analyzed that factors prohibiting participation in the industrial revolution of some states were due to internal ‘forces of inertia’ rather than a lack of cooperation between nations (Graff, Kenwood, & Lougheed, 2014, S. 129). While that provides an optimistic outlook for technology-appreciating states, politics at current does seem to lean towards more protectionist policies; possibly comparable with the early 1900s or even the 1930s when the aftermath of the Great Depression led to social disruptions which further prohibited economic recovery and wealth creation. 1914-1945 is generally considered a time of deglobalization and depression (Broadberry & O‘ Rourke, 2010)where some of the wealth created before the first world war had been destroyed.
It remains to conclude that a prediction for economic performance and therefore employment, following Okun’s law, is ambiguous at best, given certain social movements against globalization and further technological progress, for those two factors were the main drivers of employment during the industrial revolution. If, however, the trust in international trade, labor mobility and technological innovation exceeds contemporary protectionism, automation can be seen as being a positive contributor in fostering economic growth and supporting employment.
Autor, D. H. (2014). Skills, Education, and the Rise of Earnings Inequality among the ‚Other 99 Percent.‘. Science, pp. 843-51.
Broadberry, S., & O‘ Rourke, K. H. (2010). The Cambridge economic history of modern europe Volume 2.Cambridge: Cambridge University Press.
Cognizant. (2019). Globalization & the Fourth Industrial Revolution. Retrieved from Cognizant: https://www.cognizant.com/world-economic-forum-highlights-2019
Frey, C. B., & Osborne, M. A. (2017, January). The future of employment: How susceptible are jobs to computerisation? Elsevier Inc., 114, pp. 254-80.
Graff, M., Kenwood, A. G., & Lougheed, A. L. (2014). Growth of the international economy, 1820-2015(5 ed.). London: Routledge.