Go back


06 / 02 / 2020 Carlos Lastra-Anadon


In this video, Professor Carlos Lastra will answer all your questions about how automation is transforming our lives—especially in the context of the job market. While it’s true that automation will make a huge impact on employment, especially for the middle class, this doesn’t necessarily mean that jobs will be lost. On the contrary, we should expect to see more job creation.


Professor’s bio

Carlos Lastra-Anadon is an Assistant Professor at IE School of Global and Public Affairs and the Director of Research at the Center for the Governance of Change. He recently completed his PhD in Government and Social Policy at Harvard University and was a postdoctoral research fellow at The Europe Center at the Freeman Spogli Institute at Stanford.

His research interests lie at the intersection of political economy and policy, particularly education policy.



Technology shocks, employment and political changes

We all know that technology has transformed and is continuing to transform much of what was familiar to us. In fact, this accelerating change is perhaps the defining feature of our times and explains many of the other cultural and political transformations around us.

Most directly, it affects the jobs that are available, which have changed dramatically. Entire industries have been created or wholly transformed, jobs have disappeared and new jobs have emerged.

So how will this affect jobs in the future?

It is undeniable that the automation of many tasks will transform the jobs that there will be available and some estimates are that half of jobs will disappear as a result. Others go much further and think that AI or deep AI will essentially erase most jobs, if not all –except for those of the programmers, designers or owners of the robots. Algorithms and robots will be able to mimic every task, including those that have been hardest to automate from taking care of the elderly and the children to build all infrastructure, as well as all the current white collar jobs.

Beyond the hype and fearmongering, what does the empirical evidence say?

The first thing to note is that at some level, all changes so far have been part of processes of transformationsimilar to previous historical periods that have not had catastrophic effects. For instance, In the United States, 41% of all jobs were in agriculture in 1900 and by 2000 they are 2%. Yet unemployment is at the lowest it has ever been.

Arguably, the latest wave of technological change resembles this: research suggests that especially big technological advances generate many new jobs. As an illustration, there are more retail banking jobs now than when ATMs where first introduced in the 1970s: cost efficiencies generated by automation of routine processes created opportunities for greater investments in sales teams, etc.

Secondly, however, it is clear that there are challenges too. The impact of automation on the labor force differs by socioeconomic level and will be concentrated in certain geographies. In what we know so far, from research of David Autor and David Dorn as well as others, there has been a U-shaped impact since the 1980s, whereby it has been middle-skilled professional jobs have been eroded. This has meant profiles such as machine operators in factories of all kinds or clerks have virtually disappeared. By contrast there are more managerial level jobs and more low-skilled jobs (such as many in the hospitality industry or health-aides), and they earn more than they used to.

Naturally, a lot of these jobs are concentrated in industrial areas –now post-industrial. Since 1990 to the most recent data (2015) employment in industry has been reduced by a third in Western European countries, going for example from 36 to 27 of total employment in Germany, 25 to 20 (US), 32 to 19 (Spain).

What will be the consequences of this transformation?

These uneven fates for different groups promise to create ever growing backlashes. It is not hard to reach the conclusion that a large part of turmoil in politics we are experiencing, from Trump to Brexit to waves of nationalism across countries is related to the dislocation of traditional industries. As an example, in a survey we did this year at IE Center for the Governance of Change, we see that three quarters of Europeans are in favor of very robust government interventions (more so than any on the public domain), such as forbidding the introduction of more machines if they will displace workers, introducing special taxes on companies that replace workers or outright banning the use of machines except for dangerous or unhealthy tasks.

It is not all bad news. We know some of what can be done.

In my own research, with collaborators at Stanford and New York University, we have established that historicinvestments in higher education mitigate the negative regional effects that having high levels of exposure to routine occupations in the 1970s has on the share of population employed. Education, especially investments in community colleges, (tertiary (post 16) vocational education really pays off in terms of mitigating the negative effects of automation on employment.

Moreover, we also show that these investments have the effect of moderating the politics of these disruptive transformations. We show that for US counties that spend relatively more in higher education, this spending causes them to swing less towards voting for anti-government candidates (such as Donald Trump) from the 1980s to the mid 2010s. We also show that in those counties, citizens are more supportive of government spending in higher education and overall, presumably because they see how effective government intervention can be in mitigating negative outcomes.

In summary, it is undeniable that there will be many challenges ahead for jobs, and that manyin the developed world, especially the middle classes have been challenged and continue to be challenged. But we know at least that government spending, if well targeted, such as in investments that can directly create opportunities for citizens has been effective. Through that effectiveness in creating opportunity it can serve to moderate some of the positions that politics has taken in recent times.