Pascual Restrepo on the economic impact of automation on global labor markets
by Adam Walker
According to projections by the Future of Jobs Report 2020, by next year 85 million jobs globally will have been displaced due to shifts in the division of labor between humans and machines. Automation is increasingly taking over tasks traditionally carried out by humans, and the rapid emergence of artificial intelligence risks even broader displacement. But while much of the discussion around automation has focused on high-income countries, its impact on emerging economies remains less explored.
New research by Pascual Restrepo, an Associate Professor of Economics at Yale and an EGC affiliate, offers insights into the future of work and inequality through his research on how automation affects labor markets. Technological advancements in automation can boost productivity and improve living standards; however, they also present significant challenges by displacing workers. Restrepo’s work investigates how these shifts contribute to growing income and wealth inequality, particularly as automation becomes a more dominant force in labor markets worldwide.
Building on existing literature, Restrepo’s research examines the uneven impact of automation. He warns that even without widespread automation within emerging economies, they could still feel the effects as multinational companies relocate operations back to developed countries where automation may replace the need for cheap labor.
Shifting Gears: How Colombia’s challenges led Restrepo to automation
“I've always liked mathematics and economics – that was my thing since I was a kid,” Restrepo said in an EGC interview. Born and raised in Medellín, Colombia, he studied these subjects as an undergraduate, eventually focusing on Colombia’s economic paradoxes, like the coexistence of strong democratic institutions amid persistent conflict and economic instability and the illegal drug trade’s role in fueling conflict and violence.
“Colombia is very interesting, in terms of development,” he said, noting that Colombia has maintained stable macroeconomic policies, benefited from an independent central bank, and steered clear of extreme political regimes. Nonetheless, the country has grappled with drug trade-related violence and continues to face difficulties in strengthening state capacity across much of its territory.
As Restrepo pursued his PhD at the Massachusetts Institute of Technology, his academic interests evolved from conflict economics to the effects of automation on labor markets, inspiring him to explore the long-term impacts of technological change on income distribution and wealth concentration. He was struck by how automation was reshaping work in both developed and emerging economies and how global labor demand was changing, with new technologies replacing traditionally labor-intensive jobs. This realization led him to investigate how technological advancements could drive both economic growth and inequality, especially in countries already facing development challenges, like Colombia.
After receiving his PhD, Restrepo joined Yale’s Cowles Foundation as a postdoctoral fellow, where he was able to dive deeper into research on automation. “It was a year where I didn’t have to teach, so I could just focus on research,” Restrepo noted. “That gave me the space to push forward on the big questions I was asking about technology and inequality.”
The uneven distribution of automation’s benefits
“Automation is one of the ways in which technology allows us to produce more, but it’s particularly relevant because it works by rendering human labor obsolete in some areas of the economy – not every area, but some,” Restrepo said. “That’s the main focus of my research: how does automation affect countries or people, depending on whether they were the ones doing the thing that got automated?”
These advancements, while driving productivity, also lead to what economists call “general equilibrium” effects: widespread automation in one sector can have cascading economic impacts, affecting wages but also market dynamics.
Restrepo’s structural analysis of global automation trends reveals that the benefits of automation are not equally distributed. For instance, large capital owners, particularly in developed countries, are best positioned to capitalize on the gains from automation. This is because automation tends to be capital-intensive, requiring significant investments in technology and infrastructure. As a result, the wealth generated by automation often accrues to those who control capital rather than those who rely on labor for their livelihoods.
In lower-income countries, where labor remains abundant and cheap, automation has less immediate relevance. Yet Restrepo warns that the slow adoption of automation in these regions may result in lost opportunities for growth, as multinational corporations shift their investments toward fully automated production facilities in developed nations.
Restrepo’s research highlights the global implications of automation. In the past, multinational corporations have sought out developing countries with abundant and cheap labor to set up manufacturing operations. With the rise of fully automated production systems, the need for cheap labor is diminishing. Restrepo argues that this shift could begin to reverse decades of economic growth in developing nations, as companies begin to favor automated production in developed countries, which also offer more stable institutions and lower transportation costs.
In collaboration with Daron Acemoglu, Professor of Economics at MIT and recipient of the 2024 Nobel Prize in economics, Restrepo explored how automation affects inequality and productivity. They showed that while automation has been key in shaping economic benefits – often widening gaps between demographic groups, such as those with varying levels of education – it has not substantially boosted overall productivity. They argue that simply automating tasks previously performed by workers is not enough to drive sustainable economic growth while avoiding increased inequality. Instead, they emphasize the importance of creating new types of work that leverage human skills to support job growth, wage increases, and a more balanced distribution of resources. They also introduce the concept of “machine usefulness” over “machine intelligence," suggesting that technology should focus on enhancing human capabilities rather than solely replacing human labor.
Restrepo’s collaboration with Acemoglu has been personally rewarding – for both economists.
“Pascual is a force of nature,” Acemoglu said. “I have learned so much from him both on theoretical matters and on empirical work. He was my student, but over the years he has become my teacher. His blend of creativity, attention to detail, seamless ability to combine many different approaches and a great knack for sensible assessment of the real-world lessons from models and data is simply unique. To top it all off, Pascual warms the heart of everybody he comes into contact with.”
The future of automation and policy implications
As automation continues to reshape global economies, Restrepo’s research emphasizes the need for targeted policy interventions. While some have proposed solutions like universal basic income (UBI) to counteract these trends, Restrepo is skeptical of its effectiveness on a global scale. He points out that UBI may address inequality within individual countries, but it does little to solve the broader, international wealth disparities exacerbated by automation. He also stresses the importance of international cooperation to address the widening global inequalities created by automation.
In his current role at EGC, Restrepo is continuing to explore these complex dynamics. “I already know the department because I was here before, so it’s like coming back home,” he said of returning to Yale and joining EGC. “I’m excited to continue pushing forward on automation research, especially with the incredible resources and colleagues here at Yale.”