US Videos

Automation Isn't the End of Jobs

Christine Benz

Note: This video is one of several interviews that Morningstar director of personal finance Christine Benz had with Vanguard officials at this year's Bogleheads event. See all of the interviews here.

Christine Benz: Hi, I'm Christine Benz for Morningstar.com. What will the increase in automation mean for the future of jobs in the economy? Joining me to discuss some recent research on that topic is Joe Davis. He is the global chief economist and global head of Vanguard Investment Strategy Group.

Joe, thank you so much for joining me.

Joe Davis: Thank you, Christine.

Benz: Joe, let's talk about this future of work paper. You examine trends in employment and obviously, one big topic in the employment space has been that jobs are changing and some going away due to automation. The starting point for your research was that you decomposed jobs into a series of tasks and you noted that some of those tasks are more susceptible to automation than others. Let's talk about that overall idea and why that was kind of the genesis.

Davis: Thank you, Christine. More than two years ago, what got us really focus on this research, there was a number of eye-opening studies from academia some of which were arguing that as much as 50% of the U.S. jobs in the foreseeable future would disappear, which obviously, if we think about the markets and the economy, that would have significant ramifications. That certainly got our attention.

However, we did the analysis ourselves, and what we found is, I think, some of those studies are too pessimistic because they were characterizing the threat of automation solely through the lens of a job, when I think much more importantly, every job is a distillation of different tasks. The average job or a profession--say, an economist, for example, has 30 or 40 different tasks. It could be everything from processing information to solving a problem. When you do that, there are clearly occupations that will have a higher probability of automation going forward. But the general conclusion is, 10 years from now we will likely, very likely, have more jobs in the future, not less, although the composition of jobs is changing. 

Importantly, the unique insight, I believe, that our research brings upon that debate is that it's not necessary thinking of it as one occupation versus another occupation. All the jobs are changing significantly in the labor force regardless of one's industry.

Benz: You classified the tasks into three major groups: basic, repetitive, and what you call, human or uniquely human. Can you talk about what that is and also give an example of each task within those categories?

Davis: Almost anyone--regardless of what someone's profession may be, everyone likely has some sort of percentage attributed to one of these three components. There's roughly 50, we characterize 50 rough categories of tasks into those three broad buckets. An example of a basic task: usually, tasks associated with vital human survival. That could be everything from growing food to also just accessing information. It's not just manual work. Repetitive tasks: things such as assembling, something that can be effectively codified by the computer. Repetitive tasks are most likely the ones that would be automated a percentage of time.

Encouragingly, the biggest and largest bucket is uniquely human task. They involve skills and tasks such as things related to creative intelligence, emotional intelligence, things such as solving problems, coming up with creative solutions, leading teams, managing others. Professions that have a high percentage and uniquely human tasks, not a 100%, is wide-ranging. It's everything from certain cooking professions to nursing to various managerial professions and even certain parts of the manufacturing sector.

That's where, I think, this debate of only certain occupations are going to have job growth; when you look through the lens of task, it actually says there's a much broader base for potential for job growth going forward as well as more disruption in the change in our own lives. Because again, what we found just in the past 15 years, 50% of our time has changed between repetitive and uniquely human tasks. That's good. I think the labor force, as we speak, is going through a significant change. Those studies assumed that professions don't change which is why they bet that most jobs will go away, and that's the critical error, in our opinion, that they make. Jobs are changing as we speak, which means it will likely become more efficient and more valuable over time.

Benz: You mentioned, one of the data points that really jumped out at me--U.S. labor force spent 30% of its time on uniquely human tasks in the year 2000; 15 years later, it was 50%. You pointed out in the paper that the global financial crisis hastened that. Let's talk about that and what was going on.

Davis: It really accelerated. I think part of this is natural, just with the drive for companies on their profitability duress to really become more efficient. We've seen this throughout history. The global financial crisis wasn't the only time where we had a push towards more automation as well as, unfortunately, the firing and letting go of some workers. But what we are finding is, yes, it did accelerate it to some extent but it's not a surprise that on a trend basis there are certain percentage of our jobs and tasks where our work becomes more valuable with the assistance of the computer. There's other tasks that can be substituted away from the computer. We're seeing, across various professions, both the percentage of time as well as the new jobs being created, much more toward the uniquely human component where computer is an asset to one's job rather than doing the work of a computer. So, processing information, accessing information--that process was already in place. It just accelerated a little bit. I don't see that acceleration slowing down anytime soon.

Benz: Another point that you went over in the paper was geographic dispersion of the uptake of these uniquely human tasks. I think one might think, well, they are highly concentrated in the really highly educated more affluent positions. Not necessarily. You found geographic dispersion.

Davis: That was a surprise to me. I didn't expect to see it, Christine. What we're finding is that this radical change in the type of work that everyone is doing is not just centered in the high-tech centers. Although, clearly, they have had change and are benefiting. It's almost every county in the United States. More than 80% of the counties in the United States have seen a significant increase in the percentage of time spent on uniquely human tasks. Even if one is in the same job that one was in a decade ago--think about even how you and I do our jobs versus a decade ago, which gives some hope that there's still an ability and potential for workers to upscale. Now, that need to upscale was not going to change. It is going to continue to be there. But that was a positive surprise for us.

We also found that jobs regardless of the income distribution, meaning the average wage in those professions, are generally seeing a percentage of time spending more uniquely human. So, this is not an argument of lower income versus high income, which is generally how it's portrayed as well as Silicon Valley versus everyone else. That's too rough a characterization.

Benz: I guess, another interesting aspect of this is that those uniquely human tasks might be sort of fundamentally more satisfying for us, right?

Davis: The one thing that changed my mindset with this research, Christine, is that if I had to pick three skills--so, again, I'm not an educator, but I would say in regards to if I'm in the parent, as a manager of a workforce, as an employee myself--there's three critical skills going forward. One is around creative intelligence. That critical need is there.

Benz: You want to nurture that in your kids …

Davis: Nurture that. Emotional intelligence, which is really critical, because that's the human component to it, everything from leadership to management. Even if one is a salesperson, is more of an individual contributor, that's going to be increasingly important. The third is technology acumen. That does not mean--and I'm not here to disparage in any way, shape or form the drive towards STEM fields and so forth--but we have not seen material changes in the number of STEM jobs over the past decade. They will be a small percentage of the jobs created going forward, although they are very important fields. Technology acumen is going to have to increase for a whole swath of the population. We are seeing that change occurring, but my only worry going forward is that the speed of change is going to outpace the ability for individuals to change. I think, ultimately, we will weather through that, but there's a profound change going on in the labor force which will continue.

Benz: This is obviously a huge trend. I think you call it a megatrend. In terms its impact on your overall economic outlook, how does something like this affect that?

Davis: It's a fair question, Christine. What this has profoundly led us to for the past two years is to say because we came to the conclusion there was going to be more jobs in 10 years, not fewer and an increased demand for more uniquely human professions and tasks and that demand for those jobs and those tasks will outstrip the supply of them. That told us that the labor market will continue to tighten more quickly than the market was anticipating. 

In other words, the unemployment rate, if the forecast two years ago was for it to stop at 4.5%, we said based upon analysis it's going to be lower. Here we are today 3.7% and it could be below 3% in a year or two. The irony is that we have a great deal of automation. Those headlines will not go away, but the tightness in the labor market will be tighter than we've had in the past. That's influenced our thinking in terms of the Federal Reserve policy. It means that short-term interest rates over the next five or 10 years could be a little bit stronger or higher than we would have previously thought, all from doing research on automation. That's the implication.

Benz: Fascinating research, Joe. Thank you so much for being here to discuss it.

Davis: Thank you.

Benz: Thanks for watching. I'm Christine Benz for Morningstar.com.