Economic Mobility Project - The Automation of Jobs : Impacts on Workers and Inequality

ANNA PAULSON: Good afternoon, everyone, and thanks for your patience as we worked out a few technical issues. I'm Anna Paulson, Executive Vice President and Director of Research here at the Chicago Fed. Thank you so much for joining us for the Automation of Jobs, Impacts on Workers, and Inequality. The COVID-19 pandemic has significantly impacted the American labor force, accelerating automation and exacerbating pre-existing economic inequities.

Today you will hear two impactful presentations from our economists. In the first, Kristin Butcher, Chicago Fed Vice President and Director of Microeconomic Research, will discuss the lasting negative effects of job displacement on workers with different demographic and socioeconomic characteristics.

Next, you will hear from Kristen Broady, Director of our Economic Mobility Project. She will discuss the impact of automation on workers during the COVID-19 pandemic. These presentations will be followed by a roundtable discussion with experts in the fields of labor, automation, and higher education.

We're very pleased that you've joined us this afternoon. And now I would like to turn it over to Kristin Butcher. Kristin, the floor is yours.

KRISTIN BUTCHER: Thank you so much. Can you hear and see me? Excellent. So, Peter, can we go ahead and start sharing those slides? All right. Thank you so much. Today I'm going to talk about job displacement in the United States. And we're going to look at the effect of job displacement by race, education, and parental income. This is joint work with Ariel Gelrud Shiro, and it was undertaken when we were both at Brookings under the auspices of the Future of the Middle Class Initiative and the Center on Children and Families. Next slide, please.

So let me tell you what I'm going to tell you before I tell it to you. There's a long literature on the impact of job displacement, and it all shows that job displacement is not good for labor market outcomes, and it has both immediate and long term impacts. People have built on that literature and showed that there are adverse impacts on mortality on the children of the displaced workers. So this is something that has both immediate and very long run consequences.

The recent literature is focused on heterogeneous effects. So are there differences between groups of the effect of job displacement? Here, as I said before, we're going to talk about those differences by race, education, and parental income. What we're going to find is that for all groups, there's both an immediate and a long term impact on annual earnings, and that these effects are very similar across groups. However, the likelihood of being displaced is higher for less privileged groups. Next slide, please.

So what do we mean when we talk about job displacement? Well, we're really thinking about somebody's job leaving them, not them leaving their job. And we're particularly concerned about times when somebody may have built up some specific skills that help them be particularly productive on that job. What are some sources of job displacement? Technological change, regulation, trade, and if your job leaves you because of one of these reasons, it's very unlikely that you're going to go out and find a very similar job.

And so those investments and skills that both the firm may have made in you and that you made in your firm, those are not going to be available to you anymore. So there's going to be a real loss here. As I said before, we're really standing on the shoulders of giants here. The first, the canonical investigation of this is Jacobson, LaLonde and Sullivan back in 1993, and I'm going to show you some results from that work. They also built on that work to find that there was impacts on mortality.

Oreopoulos, Page, and Huff Stevens have found that the children of displaced workers suffer worse outcomes compared to those children whose parents are never displaced. And more recent work by Kaila, Nix, and Riukula uses really detailed Finnish data from Finland to look at the impact of job displacement based on parental income. And they find that people who have wealthier parents suffer less when they get displaced from their jobs.

And then there's another paper using the same data that we use that looks at parental proximity. And that work finds that people who live closer to their parents seem to be buffered when there's a displacement. So we're building on this early literature, and we're placing ourselves in this newer work that's looking at the heterogeneous impacts. Next slide, please.

So this from the canonical, Jacobson, LaLonde and Sullivan, really sets up what we're talking about. This was really innovative work. It's using administrative data from the state of Pennsylvania. And what we're looking at is high attachment workers, people who have 6 or more years with a firm. And then that top hashy kind of line is showing you what happens to the earnings of workers who are never displaced. And then the bottom line is the workers who are displaced in that first quarter of 1982.

And this is really the name of the game in all of the literature. It's trying to find someone to benchmark against so that you can say, well, what would have happened to your earnings if your job had never gone away? And what we can see here is there's this immediate drop, and then earnings recover somewhat over time, but never go back up. So it really looks like it's a permanent loss of those skills that were built up during this substantial period of time that the workers were with the firm. Next slide, please.

Everything that you look at in this literature is going to look a bit like this. So this is just taking that picture that we were just looking at, and it's graphing the gap between the workers who were never displaced and the workers who were displaced.

And it's lining that up around this 0 mark so that we're not limited to just look at people who lost jobs in the first quarter of 1982, but at whichever point you lose your job, we're just going to call that 0, and we're going to look at your 5 years before. And in this case, 6 years after. And again, everything we're going to see has this sort of flavor.

The key to this is that we can see that 3, 4, 5 years before the displacement occurs, these workers are very similar to the workers who are not displaced, suggesting it's not something about the workers that's causing the displacement. We see a drop off leading into the displacement suggesting that if you're working for a firm that's going to go away, maybe you're not earning quite as much. But the real impact comes when you lose your job, then there's a rebound, and you never quite recover the value of those skills that were firm specific. Next slide, please.

Now, this is turning to our work. This is using a very different data set. It's using the panel study of income dynamics. So here we're not relying on administrative records, but rather, what people say about themselves.

And we're going to use a different definition of substantial time with the employer. We're going to look at people who have 2 or more years of full time work with their employer. And there's a question on the survey where they ask people who've recently changed jobs, what happened to your last job? And we are designating those people who say, the plant closed or the employer moved. We're calling those people displaced.

And what you can see is something very similar from what we've seen in the other pictures. That solid line is the average difference in earnings between the displaced and non displaced workers, and the shaded area just shows you something about the statistical precision.

And so, what we can see is before that displacement, that shaded area overlaps 0. So there's really no difference in the earnings between the displaced and non displaced. And there's a ginormous-- that's the technical term-- impact of displacement on annual earnings, and then it increases, but never goes back up. So again, in this data set, which is really quite different, we see a very similar pattern. Next slide, please.

So, next, we're going to turn to looking at differences by groups. And the groups I'm going to look at are Black workers versus white non-Hispanic workers, workers with a 4-year college degree and no 4-year college degree, and then workers who have parents whose income is above the median and below the median for parental income. You might ask, why might these things be different?

Well, if you think that some groups of workers never have the opportunity to be invested in or to invest in an employer, or they have the types of jobs like a cashier that you could just move off to another employer who needs cashiers and you wouldn't lose the value of that particular type of skills you've invested in, that particular type of work, then maybe you'd expect there to be very little impact. It could be that some workers have better connections than other workers and are more readily able to re-establish themselves after this kind of shock. Or some workers might have the kind of broad-based training and skills that make them resilient to shocks.

So there are reasons we might expect there to be differences in the impact by group. Next slide, please. In fact, we find that there really aren't substantial differences. This is between Black and white non-Hispanic workers. The Black workers are in the orange and the white workers are in the dark blue.

And what we can see is that the estimates are noisier for the Black workers as there are fewer of them in the data set, but these overlap substantially. So everybody is suffering an immediate loss when they're displaced from their jobs. And then they're recovering some, but typically not all the way back to their pre-displacement earnings. Next slide, please.

I'll just quickly go through it for the other groups because it's a very similar story. Those with a BA versus no BA also quite similar. Next slide, please. Those with parents who are higher versus lower income, also very similar. So everybody's losing. Everybody's losing for a long time, but no really big differences between groups. Next slide, please.

But here's what we do see differences between groups-- the risk of displacement is higher for less privileged groups. So for Black workers, that risk is 2.5%. That's the share displaced in these data. For white workers, it's only 1.5%.

For those with no bachelor's degree, it's quite a bit higher. For those with lower income parents, it's quite a bit higher. So although everybody suffers from this type of job loss, job shock, there are some groups that are much more likely to suffer from that. And we're going to talk in the next presentation about the risk of automation for Black workers. Next slide, please.

So what we've seen here is that there's both an immediate and a long term impact, but the likelihood of being displaced is much higher for less privileged groups. And it's important to keep in mind that this type of displacement increases during recessions and that technological change can increase displacement. I don't think we want to be in a world where we're against technological change, so I think it's incumbent upon us to think very carefully about what kinds of policies can help workers who are at high likelihood of this type of displacement.

And just an example, traditional unemployment insurance is really ill-suited to support displaced workers because the model of that type of support is very short term. The jobs went away, let's just wait, let's just support your income until demand resumes and you'll get the same kind of job. And what we're seeing here is that's not what's happening when your job leaves you and doesn't come back from something like technological change.

Another area I think we need to investigate are training programs-- both training programs that might prepare people prior to displacement so that they're less likely to be displaced or they're more resilient upon displacement, but also I think we need to think hard about what are appropriate training programs for people who've suffered a displacement? The bet they made on the skills that they need to lead a long and happy life just didn't turn out to be right, and what can we do for them?

All right. With that, let me turn it over to Kristen Broady. Thank you so much.

KRISTEN BROADY: Thank you, Kristin. Hopefully everyone can hear me. So I am going to now talk about jobs at high and low risk of being automated. My presentation is based on research that I conducted with Darlene Booth-Bell, Anthony Barr, and Ryan Perry here at the Chicago Fed.

So hopefully Peter can start to share my slides. I'm not sure if everyone is able to see them here. Here we go. All right. There we go. So the title of the presentation is The COVID-19 Pandemic Spurred Growth in Automation: What Does This Mean for Minority Workers? Next slide, please, Peter.

Saturday, January 21, 2023 marked the 3rd anniversary of the first confirmed case of COVID-19 in the US. Now in the 4th year of the pandemic, the World Health Organization reported on February 21st that there had been 757 million confirmed cases of COVID-19 and 6.8 million deaths.

More than 1.1 million people in the US alone have died from COVID-19 related causes. The contagious nature of the virus necessitated physical distancing and remote work for millions of workers. And employment fell by 20.5 million jobs between March and April 2020, by far, the largest 1 month decline in Bureau of Labor Statistics data collection history. And the corresponding unemployment rate increased from 4.4% to 14.7%.

On December 11, 2020, the FDA issued an emergency use authorization for a two-dose COVID-19 vaccine. A year, later 72% of Americans had received their first dose of the COVID-19 vaccine, and 61% had received two doses, completing the series. As vaccines became more widely available, companies accelerated their reopening plans, while many workers had transitioned to telework, who were teleworking, were not ready to return to their workplaces. They were not ready to go back into those offices.

64% of workers who were teleworking said that they would feel uncomfortable returning, while only 31% said that they would feel comfortable. During what has become known as the Great Recession, but I think of as a great transition, 47.8 million people quit their jobs in 2021, comprising 69.3% of total separations. Layoffs and discharges decreased by 23.8% million to 17 million, comprising 24.6% of total separations. Next slide, please.

So I want to talk to you about jobs at high risk of automation. Frey and Osborne, 2017, examined the expected impacts of future computerization on US labor market outcomes. Drawing from a workshop held at the Oxford University Engineering Sciences Department, they examined the automatability, that is, the ability of a job task to be completed by a computer or a computerized technology of a range of tasks associated with job descriptions for occupations, and answering the question, can this task be completed by a computer or a machine?

And so they ranked the occupations according to their probability of computerization from lowest to highest. They considered occupations with an automation probability of 70 to 99% as high risk of being automated. Next slide, please. So now I'd like to show you the subset of 30 jobs with the highest automation risk scores that employ the highest number of US workers. And here we start with the first 15.

And so we see that the first job is retail salespersons, which employed 2.75 million Americans in 2021. The job had an automation risk score of 92%, an average hourly wage of $18.61. And employed in this job were 48% female, 78.3% white, 13% Black, 4.8% Asian, and 20.8% Hispanic. And so this slide lays out these top 15 jobs and shows the demographics.

And so you can see some of the other jobs-- cashiers, construction laborers, secretaries, cooks-- all of these jobs, many of the tasks can be done by a computer or machine right now, and certainly in the near future. More and more of the tasks associated with these jobs are being able to be completed by some type of computer or smart machine. Next slide, please.

And so as we finish out the list, we see that these jobs employed 31 million workers, had an average automation risk score of 92.67%, an average hourly wage of $21.21, which sounds pretty good when we think about the push for $15. That doesn't sound like such a low wage when we think about it. And so, in these jobs, we see that they were 50.6% female, 77.7% white, 13.7% Black, 4.4% Asian, and 25.1% Hispanic. Next slide, please.

Now, Black and Hispanic workers are over-represented in these jobs at high risk of being automated. Black workers were over-represented in 17 of the 30 jobs and in this group. They represent 13.7% of the employment in these highly automatable occupations compared to 12.3% of the civilian workforce. Hispanic workers are also over-represented. They account for 22 of these jobs, right? And so they are 25.1% of these jobs, but only 18.7% of the civilian labor force. In addition, women accounted for 50.6% of the employment in these 30 highly automatable jobs while they accounted for 47% of the total labor force. Next slide.

So now I want to talk about jobs at low risk of automation. And one of my favorite economists who you're going to hear from shortly, David Autor, would classify some of these jobs as being in the knowledge class. People who think for a living, who are more likely to be able to telework, who earn higher wages on average. And Frey and Osborne classify these jobs as being at very low risk of automation. So let's see what they are on the next slide.

And so as we start, we see that the number 1 job, elementary and middle school teachers, employed 3.26 million people, had a very, very low automation risk score, 0.44. So we're going to have teachers for a very long time. It would be very difficult for a computer or a machine to not only teach children, but keep them engaged and conduct any discipline that would need to be done. And so we see that the average hourly wage here is $29.48. And we see that, when we look at the demographics, teachers where 79.2% female, 83.8% white, 10% Black, 3.4% Asian, and 10.7% Hispanic.

So as we continue on, the numbers don't look too different, right? The first two. The second one, registered nurses, more than 3 million people employed. We do see very, very, very low automation risk scores. And we're starting to see higher wages as well. If you look at physicians, $101 an hour. So let's go on and look at the rest of these jobs from the next slide. Here we go. So we can see that these jobs-- now again, for the high risk jobs, there were 31 million people. Here we see 19 million, so far fewer people are employed in these low risk jobs.

We see that the average automation score was just 0.9. Not 92%, but 0.9. So not a whole lot of the tasks in these jobs are going to be able to be done by a computer or a machine anytime soon. And that $21 an hour average hourly wage from before may have sounded good for those high risk jobs, but not compared to $46.76 for these low risk jobs. So that's more than double. And we see here that we have 52.2% of the female workforce, 80.6% white. And so that's an over-representation. 9.7% Black, under-representation. 7.3% Asian, over-representation, and just 10% for Hispanic workers. So let's go to the next slide.

So we see here that white workers were over-represented in 22 of these jobs. Black workers are over-represented in just 10 of these jobs. Asian workers were over-represented in 12, and unfortunately, Hispanic workers are not over-represented in any of these jobs. This is the third version of this paper. We worked on this using 2016 data, 2019 data, as well. And still, in 2021, we do not see an over-representation of Hispanic workers in any of the jobs at low risk of being automated. Next slide, please.

The pandemic has influenced a variety of intersecting trends around consumer behavior, labor markets, and the degree of automation. First, the COVID-19 pandemic drove increased e-commerce activity. The shift to online shopping increased revenues in online marketplaces.

As an example, in the first half of 2020, revenues at Amazon grew 34% compared to the same time in 2019, and by 27% for Alibaba, and 74% for Shopify. The pandemic served as the impetus for the adoption in increased digitalization of workplaces with increased use of online meeting platforms, cloud computing, online contracting, and digital payment systems as examples.

For employers, the COVID-19 pandemic added what Cormack and Stiglitz refer to as a shadow cost, reflecting the dollar equivalent of all costs associated with increased risk of disease transmission on labor that requires proximity. This shadow cost, they explain, accelerated automation during the pandemic. Next slide. And so what did that mean? We're going to see some examples that Peter is going to share with us.

So in 2020, approximately 2.25 million industrial robots were in use around the world. Pandemic related labor shortages and an increasing demand for contactless delivery have accelerated the utilization of food and retail delivery robots in airports, senior living facilities, and on college campuses, among other places. Food and retail delivery robots are being used in US airports, including the Cincinnati Airport and the Philadelphia International Airport, to provide passengers with the ability to order online and receive their deliveries on a contactless basis.

Grocery stores have deployed robots to process transactions, clean floors, stack store shelves, and deliver groceries to shoppers. In 2021, 30% of grocery store transactions were completed using self-checkout machines. Now, as I share some examples, I also want to say that it's not clear whether the disparate impact of automation will translate to higher unemployment rates for various groups, or whether it will simply mean displacement into other occupations. There are a few other examples on the slide here of just how-- yes. So these various cleaning robots that we see that are doing sanitation and things like that.

So new technologies may both displace workers and generate new employment opportunities. For example, while sales from e-commerce companies like Amazon-- particularly considering the growth in e-commerce during the pandemic-- reduced the number of sales and employees in traditional retail stores, Amazon also creates new opportunities by hiring workers to work at its fulfillment centers and in other parts of its distribution network.

Mandel 2017 found that from 2007 to 2016, the general retail sector lost 51,000 jobs, which sounds like a lot. But the e-commerce sector added 355,000 jobs. So while we do see these robots that may be assisting some workers and may seem to be doing the jobs that used to be done by human workers, we are seeing people transitioning into other jobs, in many cases, jobs at lower risk of being automated. Next slide, please.

So how do we get to this future of work? How do we prepare students and workers to work with automation and not fear being replaced by it? I want to show you a few examples of what's going on out there in the world, what schools are doing to work with companies to prepare people. Next slide, please. So starting in 2014, Kentucky State University partnered with Toyota.

Incoming freshmen in this program were picked from a pool of high-achieving STEM students. They received a full ride scholarship for their first 3 years at Kentucky State to complete a pre-engineering curriculum. The last 2 years were completed at the University of Kentucky, and after the 5-year period, the students would have an undergrad degree from Kentucky State and an engineering degree from the University of Kentucky.

They would then receive employment at Kentucky Motor Manufacturing. These are jobs that are going to be at lower risk of being automated, and the students have been prepared to work with the technology that Toyota is using. Next slide, please.

The UNCF awarded Dillard University $1.2 million for-- awarded several universities $1.2 million. Dillard University was one of them. Dillard received $300,000 of that $1.2 million. And so the University introduced a new learning model. It blends technical and social skills to help students prepare for this future of work with automation. They're learning what we call soft skills and tech skills. Next slide.

And even more recently, Howard University was awarded $90 million in a contract to conduct military research for the Pentagon. Howard is going to partner with other HBCUs to conduct this research. And it's dedicated to looking at technology and AI and how that can help the military and the rest of the country as well. Next slide.

So in closing, we can't say that automation is going to displace all, or even many workers. It really depends on what companies decide to do with that technology. That determines whether a job will be displaced or whether a human being will be assisted by technology. Nevertheless, we do need to prepare for this future. And so I'm really excited to introduce our panel who is going to tell us how we get there.

I'm so excited to introduce one of my favorite economists, David Autor, who is the Ford Professor of Economics at the Massachusetts Institute of Technology, Allison Dembeck is Vice President of Education and Labor Advocacy and Government Affairs at the US Chamber of Commerce.

Juan Salgado is the Chancellor of the City Colleges of Chicago, and Thomas Hudson is the President of my graduate alma mater, thee I love, Jackson State University. We're so happy to be joined by Kate Bahn, Director of Labor Market Policy and Chief Economist at the Washington Center for Equitable Growth, who will serve as the moderator for our panel discussion. Kate, I'll turn it over to you. Thank you.

KATE BAHN: [AUDIO OUT] Kristen. And also, thank you to everyone at the Chicago Fed for having me for this really exciting panel discussion. The presentations that we saw were really illuminating and provide a great basis for us to dig into the current structures and dynamics of our economy at this stage of the COVID-19 pandemic, and then also what tools there are to ensure we have robust and broadly shared growth for workers.

So first, really quickly, I want to establish our goals, which are generally to share information on current economic research and best practices, particularly in education. Which is to say this is not about politics. I know the politics of actually passing policies is very tricky, but we want to establish a vision of what's possible here based on the best evidence and the experience of our experts.

So echoing Kristen's big question is that how do we get to the future of work? With that, I want to turn to our panelists, and I want to start with Kristen's favorite economist, David Autor. So David, what are the most significant employment transformations the American labor force has experienced as a result of the COVID-19 pandemic?

DAVID AUTOR: Thank you very much. It's an honor to be here, and I enjoyed both papers very much. The labor market has changed in very surprising ways in the 2 years since the pandemic. In particular, we've seen a very sharp reduction in inequality measured as the ratio of earnings of the 95th percentile worker to the 10th percentile worker. And a real tightening of the labor market for people, for young workers without college degrees who have seen wages mostly keeping up with inflation, unlike everyone else. They're doing a lot more job hopping.

And I think this is significant in two ways. One is, we actually knew, even going into the pandemic, before that occurred, that we were facing a tightening labor market. We have low fertility rates, dramatically restricted immigration rates, large numbers of people retiring from the labor force as the baby boom generation retires. And so there's going to be a need for young, able-bodied workers do lots of things, and not so many people entering. And so that's one thing that was already underway.

And then I think what the pandemic has taught us is that labor market competition is very healthy for workers in general, but particularly for low paid workers. And some of the low wages we see is not just there's low demand, but really, employers are competing fiercely with one another to hire those workers.

And so there's more reason for optimism at this point than I certainly expected. I was pretty pessimistic during the pandemic about what things would look like coming out of it, and now we're in better shape. But that's not a panacea. We need to invest and build skills. And I think we'll come back to this. Let me not take too much airtime now, but thank you for the question.

KATE BAHN: Yeah. So I want to follow up a little bit with you, David, and then also bring in Alison Dembeck and talk about specifically, why the pandemic itself mattered compared to other economic shocks. So, of course, there was a contagious nature of COVID-19 that required many businesses to close, at least temporarily, and required many workers to work remotely, as I am in my office at home right now.

So while the pandemic is not over, the availability of vaccines do make in-person work safer. What adjustments do you think there will be to the way people work that you think will continue into the post-pandemic economy?

DAVID AUTOR: So first of all, there were unique circumstances of the pandemic, of course. There are several reasons why I think the pandemic kind of led to this change in the way people are searching for work, low paid workers, typically. One is, of course, if workers were very sentimental about their ties to their employers, those sentiments were shattered during the pandemic when so many people were laid off.

Secondly, there was a massive increase in household savings during the pandemic because of the very successful transfer programs, which really did a huge amount to help Americans of all stripes and help us get to this rapid recovery. And so that allowed people to have money in the bank, and that makes it easier to job shop and easier to take risks. And then there's sort of the common knowledge of everyone else is doing the same. So I think that has helped to create this very fluid labor market that has people job shopping, and that makes employers have to work hard to get them.

How things are changing, well, this is one of the unfortunate ironies of the pandemic is jobs have only improved for people who had good jobs because of the work from home option. And work by Nick Bloom and Steve Davis and others really shows that workers value that. So if you're at a job you can work from home, people would be willing to give up several percentage points in earnings to have that benefit. If they're not having to give that up, they're getting that in addition. Whereas workers who do front line and in-person services, they have more disease risk, they're under more time pressure, they face more grumpy customers.

And among hotel workers, for example, now you might think, oh, people who clean hotels, oh, great. Because customers don't demand that their hotel be cleaned every day. Well what that means is when people check out at the beginning or end of a weekend, the hotels call up their cleaning staff and say, get over here. You've got to work 24/7 to do these rooms, and then don't come back again until Friday.

So material conditions of work have improved for high paid knowledge workers, as Kristen said earlier. And arguably, in some ways, been diminished for people who don't have the luxury of working from home. So wages have risen, that's good, but not everything about work has gotten better by any stretch of the imagination.

KATE BAHN: And Allison, I want to hand it over to you as well.

ALLISON DEMBECK: I think-- and Kristen mentioned it earlier, and David mentioned it at the beginning of his comments-- the pandemic in some way highlighted problems that we were already having. They weren't necessarily new problems, but it definitely exacerbated the situation. And so I think one of the things we're really seeing when you're getting out of the office worker, as David mentioned, the office worker who is now looking for more flexibility and looking for more hybrid options, there are other problems that are being encountered. I mean, we're seeing it in the child care sector, for example right?

Those were lower wage-- and it has ripple effects. Those were lower wage jobs to begin with. And maybe we can have lots of conversations about whether or not that should or shouldn't be the case, since we're talking about our little kids at a time when they're really at risk of development and we want to make sure that they're still safe and learning. But we have these jobs that, now, people don't want to do them anymore.

There's also-- I think a good thing that's happening and that people are seeing the value and the need for finding jobs that they think are fulfilling and are paying what it is that they think that they need to be earning. And so it does create some other challenges. For example, in the child care space where we need to figure out how to handle that. But I do think that there is a bit of a reckoning in and how that all plays out. Not necessarily just in terms of automation, but in jobs in general.

KATE BAHN: Great. Thank you. That's seems very complicated so far, so we'll keep digging a little bit further. And I want to bring in everyone else, too, with my next question where we'll touch a little bit on, David referred to some of the support programs that were in place in the pandemic. So the pandemic allowed particularly for expansions to unemployment insurance and a really unique moment of political possibility because it was such an immediate and big crisis.

And specifically, the CARES Act extended access to unemployment insurance to millions of Americans who were previously ineligible to receive benefits, and gave all UI beneficiaries an additional $600 per week for a limited period of time, because there was a broad recognition that these job losses were due to no fault of the workers or the sectoral adjustments over the long term that we've been talking about.

But rather, they were a result of necessary public safety measures. So how did policy interventions like the CARES Act impact displaced workers? And then what are the short and long term impacts of policies like the CARES Act on these displaced workers? And additionally, sort of final, if there's other policy interventions that you think would be most impactful, what would those look like? I know that both Juan Salgado and Thomas Hudson haven't spoken yet, so if you two want to chime in first, no pressure. But I want to open it up so we can hear from you as well.

JUAN SALGADO: Go ahead.

THOMAS HUDSON: Well, I'll start, and thank you so much for having us on this panel. When you talk about the CARES Act and from the higher ed standpoint, HEARTH dollars, it really allowed us to hang on to our workers. It allowed us to put certain things in place to allow our workers to work remote, whether it was through more digital equipment, whether it was equipping them with what they need to work from home. It had a really good impact on us in terms of retaining workers and helping them be more productive.

Going forward, we would like to see more of those tools placed in our hands, in higher education experts leaders' hands, to continue to do more of that innovative work, more of the ways that we can affect our workers in those instances. So again, it had a great effect on us. We want that type of work to continue, placing the dollars directly in our hands so we can best impact our workers.

KATE BAHN: Thank you.

JUAN SALGADO: I would just share at City Colleges of Chicago, we operate 7 colleges that serve around 55,000 students. And our students are the, if you will, in many cases, the low wage workers that are trying to up-skill to get into better occupations. And so those students of ours, those workers in our economy were the most affected by the pandemic. They were either displaced for the moment or they were working quite a bit, dealing with health conditions. So the safety net that provided to those students, I think, was those workers and those families was absolutely critical.

What we didn't see because of the pandemic was, usually, when you have a displacement like that, you have an increased skill building and people going back to college, to community college, to skills trades programs. And because those programs were very much affected and because our students are family members as well, we, like many community colleges, saw a major decrease in the pandemic of students going on to college. That has rebounded, by the way.

In this fall, quite dramatically for us here at City Colleges of Chicago, we saw a 9% increase in student enrollment in the fall. We saw a nearly 15% increase here in the spring. And so we're seeing the workers coming back for greater skills, which I think is a great sign. I believe that that safety net provided the ability for workers to stay in the game. It's not-- a spiral can occur here, Right? Where folks get so far behind, it's nearly impossible to get back to a level of stability that'll allow you to continue your education.

I believe that that's the overall impact that these CARES Act funds-- and I agree, putting them in the hands of institutions like ours, I agree with Thomas, in higher ed was a critical and important thing that the Federal government did as well.

KATE BAHN: Great. That's really good to hear, Chancellor Salgado. I want to give David and Allison a chance to talk about how these individual experiences at these two institutions reflect broader trends based on the CARES Act.

ALLISON DEMBECK: I think one of the things-- having those emergency funds was, as has already been said, was absolutely critical, both for the employees and for the employers. There were a lot of moneys built in to help small businesses stay afloat and all of that. And they were all really important.

One thing I will say, though, is that there are still money, CARES Act money that's out there that states haven't been able to spend. So when we're talking about some of the policy changes that can and should be looked at, there were, rightfully, some pretty tight restrictions put on what could be done with that money.

But because of the way various state legislatures work and the way various state governments are set up, distributing some of those funds turned out to be a little more difficult than I think the Federal government expected it to be. And so that money is tied up and can't be dispersed because, for example, of restrictions.

I'm going to come back to the child care space. Some of those child care monies can't be spent because facilities had to be already in place. Instead of, as we're seeing workers return to work, we might need new facilities to open. We should have the ability to spend some of those monies back on new facilities.

But that allows, then, people to go back to school and people to get back into the workplace. And so it does, overall, have a large impact. And so it's something that we really do need to look at, is trying to open up some of those funds so that they are able to be spent by the Federal government-- I mean by the state governments.

DAVID AUTOR: Add to that is, I hope we can learn that we don't need the moral equivalent of war to provide support to our populations in times of unemployment and job loss and household financial shocks. And we ought to be able to do that more regularly and in a more systematic way. So everyone is aware that the CARES program, especially the Paycheck Protection Program, but also the FPUC, the Federal Pandemic Unemployment Compensation Program, these were well-intentioned programs. They did a lot of good.

They had lots of administrative problems, including a lot of fraud. And that's not because the Federal government didn't care, but because it simply didn't have the infrastructure or the administrative capacity to administer benefits on that scale. And that's because of a lot of our public infrastructure and information systems are decades out of date.

The greatest example of this, or the worst example of this, is our state unemployment systems. We have 50 of them. Each one is more archaic than the last. And yet, ironically, they're all paid for mostly by the Federal government. The Federal government pays a lot of that, especially when there's a recession. It certainly did that now.

And those states all turn over their data to the Federal government after they're all done. And this would be the moment, and yet-- and I should say, during peacetime or non pandemic time, only a minority of people who should qualify for unemployment insurance benefits actually get those benefits. States make it hard for them to access them. They view it not as an insurance policy for workers but a cost for businesses and for the state. And as a consequence, these programs aren't providing the insurance function that they should.

And my strong view coming out of the pandemic is if we had a national unemployment insurance program that worked evenly across the 50 states, we would not only be able to make sure people got the benefits they needed and administered efficiently and consistently, but we would also have the infrastructure we needed to make those distributions properly when the next pandemic or national emergency arises. In other words, we wouldn't have to do it in a half hazard way that was prone to fraud and mistakes. But put that system into place now. So that would be one long term takeaway I would have from the pandemic.

KATE BAHN: Great. Thank you. And so we're bringing that long term takeaway lens, we're going to sort of move back to this longer term trajectory and talk about some of what we heard in the presentations was how job displacement has a long lasting negative effect on workers and their families.

And in particular, it may exacerbate racial, educational, and intergenerational inequality. So we heard in the presentations that Black workers are more likely than their white peers to experience job displacement, and Black and Latino workers are more likely to be employed in occupations at higher risk of being automated.

So what role has racial employment discrimination played in these inequalities? And then what policies could be implemented to assist Black and Latino workers? And this is for everyone to chime in on.

THOMAS HUDSON: Well, just historically, again, we have worked in those jobs, as you said, that are more likely to be automated, more likely to be affected by automation. So how do we bridge that gap? How do we break that cycle, if you will? Bring in more of those high income, low automation possibility jobs into the workforce and bring in more of those to African-Americans. I know as an HBCU, we pride ourselves on starting the type of programs and really emphasizing the type of programs, whether they're STEM, whether they're business. We recently started a supply chain management program, public health.

When you look at those programs that are least likely to be automated, those jobs, if you will, we typically lean towards those programs when we're emphasizing where our students should go, what they should be studying. So I think it's up to all institutions to try to take a peek into the future as far as possible and calibrate yourselves to what your students need, especially if you're servicing a minority population, such as Jackson State being an HBCU.

JUAN SALGADO: I would just add that at City Colleges, we've oriented all of our colleges to the market, to where the growth in demand is for higher wage opportunities. Some of which require transfer to some 4- year universities. Many of which don't. If you look at the list of occupations that are least likely to be displaced, we're right there. We're right there in most of all of those occupations. We're on both sides.

Our students are in the occupations that are going to probably get displaced, and our students are moving towards those other occupations. So it's high value to society, it's high value to the local economy.

One of the missing ingredients is how companies hire. We have to be frank about this. Companies often pick 10 to 12 institutions, talking about classic Big Ten or Ivy League institutions, and that's where the majority of their hiring is going towards. And so we have to shift that. We have to change that if we're going to really have an abundant opportunity, including HBCUs, including predominantly Black institutions like ours and Hispanic serving institutions like ours, in getting corporations.

We have a partnership with Aon here in Chicago, Accenture here in Chicago. These are companies that never hired from us before who are now building a talent pipeline with hundreds of new talented people that they would have never tapped in before. And so we have to multiply those kinds of efforts if we're going to really close the racial wealth gap at the end of the day.

KATE BAHN: Thank you so much. That was exactly the pivot I wanted to make, so thank you for making my job as moderator be that much easier, because I did want to ask-- the first part of the question I had is what role has racial employment discrimination played in these policies? So if Allison and David want to speak to that as well.

DAVID AUTOR: I don't consider-- I feel like I'm not the most expert person to speak on racial employment discrimination. But I think-- underscores what has already been said is barriers to entry in these jobs, many of them are educational and skills. And if we have the attitude that only people who have a 4-year college degree can have a great job, well, we're ruling out 3/4 of minority men in America. That's insane.

And employers, exactly as Juan was speaking of just a minute ago, I found that quite inspiring, our need to find a way to be the path to the sources of skill and talent that are more diverse. And that's what they should be opening doors to.

And I think employers are and will continue, in this very tight labor market, to recognize that maybe they've been too hung up on the degree instead of focusing on what skills workers come with and what additional skills they can acquire. And so, I think part of the opportunity challenge is to help employers learn how to hire talent that doesn't come with the gold plated resume. And there's lots of it. And when we do that, I think it will improve opportunity, it will actually make it easier for employers to fill the roles they need to fill. So I hope that's one direction we'll go.

Obviously, the role of our educational institutions is central to that. And I hope that technology as well-- I'm sure this is something that both of our University leaders who are speaking about it-- can be an asset in helping to modernize the way training is done and make it more accessible, more engaging, and more cost effective as well.

ALLISON DEMBECK: Just from the employer's perspective, this is something that we have been looking at since pre-pandemic. Again, it's one of those things I think that the pandemic has highlighted as a problem, but not something that the business community didn't recognize as a problem. We did. Workforce development in general, issues around racial disparities, in particular.

And so at the Chamber, the way we're set up, we actually have the US Chamber that everybody is aware of, which is an advocacy organization. But then we also have our US Chamber Foundation, and the US Chamber Foundation has done a lot of work. Our Center for Education and Workforce Development has done a lot of work around specifically a program called Talent Pipeline Management and working with our state and local chambers on how to help their members.

And then also how to help our own members, our larger members, actually develop-- I'm going to call it a supply chain of talent, but actually thinking about how to tap into not necessarily going to your Big Ten. I mean, maybe that's your local school, but look at your local area. Look at who your post-secondary education leaders are and figure out how to partner with them. It may mean having to take a class that needs to be tweaked so that it actually fits the need of what it is that you're hiring for.

But we're teaching our state chambers and our local chambers as well as our businesses how to have these conversations and how to do a better job actually in the working with institutions, working with-- placing job ads that actually list out what is being hired for versus just saying, well, you need a bachelor's degree or you need a master's degree. Maybe you don't. And so really trying to dig down into what the skills are that are needed for jobs and teaching employers the benefit of doing that. And we've had a lot of success.

And I think it's something that, especially as the labor market continues to be tight, I mean, we've got more jobs open than we have people who can fill them. Trying to figure out how to do that at a larger scale is something we've really been working on. So I'm glad to hear that some of those partners you mentioned in Chicago are ones that we have been working with. So I'm glad to hear it is working well.

KATE BAHN: Great. Thank you both. So both President Hudson and Chancellor Salgado, I want to touch on explicitly something that you both have already brought up a little bit but make it a bit more explicit. So considering what we heard in Kristen Broady's presentation about how automation is impacting various industries and occupations, what initiatives are the institutions you lead offering to prepare people for jobs at lower risk of automation?

JUAN SALGADO: I can jump in. We're on, as I said earlier, we've oriented ourselves to the areas of growth in the marketplace. Not just growth, but where growth meets a living wage. Where growth meets career opportunity. And so that's number one. Number two is, when you're in regular communication with employers backward mapping that curriculum, you can see the changes coming, whether it's in advanced manufacturing or logistics or any one of those fields that are adapting very, very quickly.

But we're also in fields like health care, and we're also in fields like education that have very little displacement. And public service as well. So we're looking at those fields, number one. Number two is, there are a lot of adults in our system. It's not just your students coming from K through 12. We've got to have a better system for adults.

We took our Federal funds during this pandemic and we put it towards something called Future Ready. We took those high demand careers that are short term, 1 year or less. Some of them as short as 1 week of training, and we made them free to any in Chicagoan.

And what did we discover? We discovered that in Black communities where populations have declined, we actually had increases in enrollment. We had increases in engagement. Black males engaging at higher levels, Black females engaging at higher levels. Latino males engaging at higher levels. And so we're going to keep that going. We call it Future Ready. And for about a $3 million in annual investment, we're able to get people into high demand occupations, adults to nudge them to go on and get the education that's going to keep them financially healthy for the long term.

THOMAS HUDSON: Yeah. And I echo my colleague in terms of moving our students to those high in-demand fields, those highly employable fields. Really, when you look at, again, those top fields, if you look at the top 50, we have space in all of them, and we're really moving our students towards those spaces. The next step, though, is connecting them with those corporations, with those employers directly so once they graduate, they have that actual opportunity to work in those spaces.

Pushing internships, pushing fellowships, those experiential learning opportunities that allow students to get a leg up in the workforce is very, very key. And that's one of the things that we've been able to do very well here at Jackson State University. I also want to go back because the point has been made but it can't be overemphasized. Our students on the front end are the ones who are typically displaced. Our students who, in terms of who work those service industry type jobs, they're the ones that were typically displaced.

The CARES dollars that we received, providing them funding for tuition, allow them to continue to access their education even when those jobs left temporarily. And some of them left permanently. So again, front end, back end. You want to be able to get them access to an education. On the back end, you want to connect them with those employers that are in those high demand fields, making sure they have an opportunity once they graduate.

KATE BAHN: Great. Thank you so much. [AUDIO OUT], I'm going to turn to you, too, and give you a chance to answer this first. And I want to note that this will be our last moderated discussion and we'll move to our audience questions after this. But for sort of a similar question for both David and Allison, so you have a chance to answer this, but going a little bit broad about what kinds of partnerships and pipelines between government agencies, corporations, and institutions of higher education would be most impactful for students and workers who might be seeking additional workforce training?

ALLISON DEMBECK: I think that that's a great follow on and was where I was going to head. There are-- this is a fantastic time to be having this conversation, because we do need to be looking at the K through 12 pipeline and the career in TechEd, what's happening in the career in TechEd space. It's all really important. And there are companies that are doing some really good work. IBM, for example, in New York has their P-Tech schools that really do create a pipeline all the way through elementary school into jobs at IBM. And that's all really important.

But Congress is starting to look at reauthorizing the Workforce Innovation and Opportunity Act, the WIOA, which is the Federal Workforce Development program. And this is a perfect time to have conversations about what do we do about incumbent worker training so that we're not having to deal with displaced employees? That we're not having all of these layoffs?

Now, granted, we're going to have populations of people who don't want to or aren't able to learn different skills for the jobs that are going to be opened. But I think it's important as we're having these conversations about automation in particular, just because automation may be replacing one particular job, it still opens up other employment opportunities elsewhere.

And so trying to make sure that we're giving people the opportunity to learn those skills so that they can be participating, it's great for employers to have access to incumbent worker training so you don't have to lay off your employees. You can say, hey, look, here's training. Partner with the school. Your job is going to change, but here's a landing place for you and here's the opportunity and here's a career growth change that can happen from that. And I think employers really want that to happen.

DAVID AUTOR: Let me add to that. I don't actually think automation risk is necessarily at the heart of all of this. So some of the jobs that Frey and Osborne list as being susceptible to automation really aren't. Like landscaping, housekeeping, cooks, and construction equipment laborers, for example, those aren't going anywhere anytime soon, in my opinion. However, I still don't want people to spend most of their careers doing landscaping and cooking and house cleaning, because those are low paid activities because they're not very specialized.

And so we ought to be targeting people at occupations that use skills, use expertise. It doesn't have to be a 4-year college degree, it just means some degree of knowledge and expertise that not everybody has and that the market values. And what could that be? Well, one example you just mentioned was health care. Health care has a lot of the relatively high paid jobs that require real skills and credential, but not necessarily require a BA or a master's degree. There's lots of opportunity there.

Similarly, there's a fantastic article in The Wall Street Journal just the other day about how America is trying to electrify everything, except doesn't have very many electricians. Right? They're all retiring. And this is true for a lot of our trades, actually, and utility work, a lot of blue collar work. It became so unglamorous that there's sort of no one in line to do it. But actually, it's going to be here for a long time. It's well paid, it's skilled, it's interesting work. If I weren't a researcher, that's what I would do.

And so moving people-- and I think this is what Chancellor Salgado and President Hudson were both saying. I apologize for calling Chancellor Salgado Juan a moment ago. He and I have known each other for a number of years. We've worked together on the MIT task force. Targeting where there are these in-demand areas where they're going to really use human skills well.

And there are a lot of them, but they're not the same ones they used to be. And some of them have to do with our changes in infrastructure, some have to do with health, some have to do with different ways of working with technology. Not all technology is substituting people. As Allison just said, some of it allows you-- it's like giving a nail gun to a roofer, right? It makes them more productive.

So and it's valuable to know what those jobs are, and, as Chancellor Salgado had said, to have an employer on the end of the pipeline-- actually president Hudson said this as well-- so that we know where that job is going. It's not enough to say, here you go. Here's a skill. Go find yourself work. We really want to assess the person, say, here's what you know, here's what you need, here's where you go. And I admire the University systems that are doing that. And I think there's real hope there. So let me pause there. Thank you so much the question.

KATE BAHN: Great. Thank you so much. And I know we're very short on time, and I want to hand it back to Kristen Broady shortly, but I do want to ask one question that someone from the audience submitted, particularly, because I think this is exactly the question that policymakers and practitioners would love an answer to. So I want to give everyone a quick chance to respond to it, but try to keep it brief if you can. So in your opinion, what is the most important metric for measuring the effectiveness of a workforce development policy in reducing inequality?

JUAN SALGADO: Mobility. Look at Raj Chetty's work. It looks at where people are-- what their earnings are pre-training and after training and looking at how that compares to how your parents did, too. So is the American dream truly alive? And looking at the populations for whom that's true and populations for whom it's not. When you look at that data, you'll see Black men in particular are not benefiting from upward mobility. We've got a lot of work to do there. And surprising to some folks in some regions of the United States, white women are not growing in their incomes and their upward mobility.

THOMAS HUDSON: I would echo upward mobility, but I'll also add alleviating under-representation. When you look at, again, the fields and when you look at the representation of those minority fields, African-American, Hispanic and Asian, will you see that level of under-representation decrease? And you start to have more of represented [AUDIO OUT] as it relates to particular [AUDIO OUT] So again, I would look at how [AUDIO OUT] within those minority demographics [AUDIO OUT].

KATE BAHN: You broke a little bit, President Hudson, but I also want to give David and Allison a last chance to answer that question as well.

DAVID AUTOR: I'll be happy to go. So the metric I would tend to use is what the OneTen Organization calls a family sustaining career, meaning an occupation where we can see there's an expectation that your wages and your skills will grow over the life cycle because you get better. You move up.

And the problem with being a cashier or a janitor or a landscaper or a cook is in many cases, there's nowhere to go from there, even if you're good at it. So that is the-- I think that's kind of the metric I would use is placing people into those jobs. They get a toehold there that's going to help a great deal.

And echoing what President Hudson just said as well, I do think that having appropriate representation in those occupations, that creates career ladders. That creates support. I mean, I'm always so impressed, our African-American and Hispanic students at MIT, they form and join organizations that the support networks that often extend outside the University to other professionals.

And I think that's great, but until you have people in place who can role model and support, it's going to be kind of a catch-22. It's hard for people to get in until people are in. So planting those seeds is going to really be consequential and beneficial. But the sooner it happens and the more it happens, the better. KATE BAHN: And last word to you, Allison.

ALLISON DEMBECK: [AUDIO OUT] that I have a good idea of what the metric itself would be, but I think everything that was discussed was really important. And we don't have equity on the Fortune 500, either, when you're looking at CEOs and board representation there. And I think just as we continue to try to be more cognizant of racial disparities, of gender disparities, of different abilities, I mean, when we talk about DEI and those issues, too, people with disabilities often also have a lot of issues with representation.

And the more people we can-- the more we can be conscious of it, I think the better we can do at actually creating equity and career mobility in the system.

KATE BAHN: Great. Thank you so much. You all brought such amazing insight and expertise, so it's been a real joy to moderate this conversation with you all. And I'll hand it back to Kristen.

KRISTEN BROADY: Thank you, Kate. And thank you to our panelists for an informative policy discussion. I'd also like to thank our audience members for joining us this afternoon. We'll be sending a post-event survey, so please be on the lookout for it. We value your feedback, and the findings will help us with future programs.

A recording of the event and a summary will be available on ChicagoFed.org/Mobility in the next few days. We'll email you to let you know when it's available. Again, thank you for joining us, and enjoy the rest of your afternoon.