Un futuro impulsado por la IA: grandes cambios en el mercado laboral

Gene Marks [00:00:00 - 00:00:09]

Do you think that the markets are tightening now, even with such low unemployment, or do you think we're going to have an ongoing labor shortage that we've been experiencing?

 

David Autor [00:00:09 - 00:00:25]

So, it depends where you're looking. So, in the lower end of labor market, people who are non-college educated doing services right, food service, cleaning, entertainment, security, something better. There remains wage pressure. Labor markets are tight. Wages have risen a lot.

 

David Autor [00:00:25 - 00:00:38]

They're not rising more. So, I don't think it's the same kind of rate of change it was back in 2022. But they've stabilized at a higher level, and they've grown much more than, significantly more than inflation has.

 

Announcer [00:00:38 to 00:00:55]

Welcome to Paychex THRIVE, a Business Podcast, where you'll hear timely insights to help you navigate marketplace dynamics and propel your business forward. Here's your host, Gene Marks.

 

Gene Marks [00:00:57 - 00:01:40]

Hey everybody, it's Gene Marks, and welcome back again to another episode of the Paychex THRIVE podcast. Thank you so much for joining us. And by the way, if you're looking for any advice or help tips to run your business, you'd like some links to our prior podcasts and other content. Sign up for our Paychex newsletter. If you go to paychecks.com/thrive, you'll be able to get lots of great information to help you run your business.

 

Now to the interview and to the podcast. I have the honor of speaking to David Autor. David is a Ford professor of economics at MIT. He specializes in labor economics, macroeconomics, political economy, public economies. We have a lot to talk about as far as the labor market is concerned. So, David, first of all, thank you so much for joining.

 

David Autor [00:01:40 - 00:01:42]

Thank you for inviting me, Gene. Pleasure to be here.

 

Gene Marks [00:01:42 - 00:01:50]

Yeah, glad to hear you. So, first of all, tell me a little bit about yourself. How did you become a professor at MIT? Tell us our audience about your background.

 

David Autor [00:01:50 - 00:02:44]

Sure. I actually had, before I went into did a PhD, I spent a number of years doing computer skills education for poor kids and adults in San Francisco, an organization that was a nonprofit. I did similar volunteer work in South Africa, and prior to that I had studied psychology and also kind of concentration in computer science. I was a sort of self-taught computer geek. So my research sort of brings together my direct engagement with individuals in their careers and their work and opportunities available, especially to people without college degrees, people who are in the lower end of the labor market, unfortunately, and how that interacts a lot with technology and changes in what tools are available, what skills are in demand, and even we also worked a bunch on globalization, how trade has also affected employment skill demands.

 

Gene Marks [00:02:44 - 00:02:46]

Fantastic. How long have you been teaching at MIT?

 

David Autor [00:02:46 - 00:02:49]

I've actually been here since 1999. This is my first academic job.

 

Gene Marks [00:02:49 - 00:02:50]

Wow.

 

David Autor [00:02:52 - 00:03:05]

It's a fabulous place. I've been incredibly fortunate. My PhD is not even in economics. It's in public policy. I sort of fell backwards into this career, so I'm one of the luckiest people in the profession.

 

Gene Marks [00:03:05 - 00:03:24]

Yeah, it's a great job. It really is. I have a really good friend who I play squash with, and he's a professor of - he actually runs the Classics department at Penn. I live in Philadelphia. And so sometimes I visit him in his office and I'm like, man, you know, you work really hard to get to that level, as you know, and I'm sure you have, but it's just a beautiful environment to be in. And the academic environment is wonderful. And the kids are great.

 

David Autor [00:03:24 - 00:03:50]

Oh, it's great. The students are fabulous. My colleagues are fabulous. And yeah, it's an incredibly interesting thing to do, and the world is constantly changing in super interesting ways. And we are in a transitional moment in many ways, in tech, technology, in demographics, in trade, in our political economy. So, so much is changing. It's really, it's exciting time to be trying to understand what's going on.

 

Gene Marks [00:03:50 - 00:04:45]

Let's talk about that a little bit. I mean, obviously, you do speak a lot about technology and its impact on the labor markets. Obviously, AI is our big, is the big topic of discussion. I write about seven times a month for Forbes, and I cover technology for Forbes, mostly for small and mid-sized businesses. And I've been writing almost exclusively about AI.

 

So, let me ask you about what you think the impact of AI in general. We're in a stage of generative AI. We're quickly moving towards AGI. That is, you know, that's artificial general intelligence. Sam Altman just said in a recent interview that he thinks it's by the end of the decade. Nvidia's CEO thinks it's within the next five years. So, that's going to have huge effects on labor markets and what jobs are. Tell me a little bit about the future. What kind of effects do you think AI is going to have on labor?

 

David Autor [00:04:45 - 00:04:54]

Sure. So, I'm going to leave the question of AGI aside. I'm a little more skeptical now. Of course, they're probably more qualified to hold their opinions than I do, but I'm going to stick with mine.

 

Gene Marks [00:04:54 - 00:04:55]

Fair enough.

 

David Autor [00:04:55 - 00:06:31]

So, let me say, how is artificial intelligence different from the technologies that precedes it? I think it's useful to paint, make the contrast. So, what we're going to call rules based traditional computing follows rules and procedures to execute programs. For something to be programmed, you need to know all the steps, because the machine, it's not going to improvise, it's not going to problem solve, it's not going to have a new idea. It's just going to do what you told it to. That's a very powerful idea, very powerful technology, but it runs into a limitation, which I call Polanyi's paradox, which the philosopher Michael Polanyi said, we know more than we can tell.

 

There are many, many things that we do. We don't know how we do them, like how to ride a bicycle, or how to tell a funny joke, or how to have a hypothesis. And for that reason, we've been unable to program them because they lean on tacit knowledge, knowledge that we understand but don't articulate. Don't know how to articulate. What AI does is it is very good at absorbing and applying tacit knowledge. It can learn from a whole bunch of unstructured cases and data. You can say, here's pictures of this thing, now identify more of the thing, without error specifying what the thing is, what makes it a thing, what makes it something, a bicycle or a chair. And so in some ways, it's kind of the inverse of traditional computing. Traditional computing is all about rules and procedures.

 

AI, in a world historical irony, is not reliable with facts and numbers. Who would think a computer technology that's not reliable with facts and numbers? It sounds like a joke, right?

 

Gene Marks [00:06:31 - 00:06:36]

In fact, AI itself is known ... I'm an accountant. It's known to be pretty bad with accounting right now.

 

David Autor [00:06:36 - 00:09:11]

Sure. Absolutely. So, what is it actually good for? Well, it's good for, it's good for many things, but let me sort of say one thing for which it is really good, is decision support is enabling people to make decisions with messy problems. And most of us deal with messy problems all the time, as professionals do. And by, say, a professional, I could mean a doctor, I could mean a lawyer, but I could also mean a contractor, an electrician, a plumber, a pilot. You have a kind of high stakes, one-off situation. How do I care for this patient? Or how do I architect this building? Or how do I remodel this house?

 

You have a lot of formal knowledge, rules and procedures, and you have a lot of examples that you've experienced, and you have to reason from that formal body of work and that set of experiences to what decision do I make right now? And there's no right answer but there's lots of wrong answers and it matters, right? There's often a life and death and decision, or it's not a question of will the building stand? That's an engineering calculation. But will anyone want to live in it? That requires. And so, AI is really good at supporting that type of decision making. It can look like, for example, AI is now used in radiology. There is no set of rules for looking at a scan and saying what is likely to be cancer versus unlikely to be cancer. It's all a lot of judgment from examples, and AI really excels at that.

 

Now we can talk about whether AI and people work together well in that setting. They actually don't at the moment, but I think as it evolves, it's going to enable, or if we use it right. Let me say AI is not an autonomous actor in this world. The future, as my friend Josh Cohen, a former colleague at MIT, now at Apple University, says, the future is not a prediction problem, it's a design problem.

 

So, if we use it well, it will enable more people to do valuable decision-making work with support from better tools. And that would be true in, you know, accounting, that would be true in design, that would be true in healthcare, that would be true in engineering, that would be true in construction, that would be true in skilled repair, right? So, lots and lots of ways it can support decision making. And it's different from traditional computing, traditional computing, like, gives you the inputs. It says, here's the number, here's the calculation, right? Now you decide what to do. This can help you with those decisions, can help guide, it, can provide guidance, like what to look at, what to think about, and guardrails of, don't do that. Don't prescribe these two drugs together, et cetera.

 

Gene Marks [00:09:11 - 00:09:19]

As a professor, how do you feel about AI in the classroom? How do you feel about your students using ChatGPT to do their work?

 

David Autor [00:09:19 - 00:10:10]

Yeah, so it's, you know, I haven't encountered this challenge much yet because the type of problem sets I assign are not very AI amenable at the moment. People are going to be using AI in the future, and I want them to be good at it. My concern is if it prevents, it stands in the way of them learning skills that they need to be able to master themselves. I don't worry about students using a calculator to do math. However, if they didn't understand that 10 times 10 is not a billion, that would be a problem.

 

It's difficult to figure out what is the foundational set of skills you need to know to operate in a sophisticated environment where you have powerful tools. There have been terrible airplane crashes caused by people flying who are inexpert when they don't have autopilot.

 

Gene Marks [00:10:10 - 00:10:10]

Sure.

 

David Autor [00:10:11 - 00:10:36]

Right? So, I worry about that. So, I think AI is going to help people be more effective in writing, in presentation, in software. On the other hand, if they don't know how to express themselves without a machine or think what visually, how to communicate, I worry about that. So, I don't think we know the right balance yet.

 

Gene Marks [00:10:36 - 00:11:27]

Right, but there will be a balance. Is there a, um, I mean, do you feel that AI is going to cause all of us, both in our professional and personal lives, to reevaluate the skills that we really need to know? I mean, if you and I were driving a car in the 1920s, we would need to have mechanical skills because cars broke down all the time and we would need. It's just natural. Now, I don't know how you are, but I don't even know how a car works. I mean, generally I do, but I break down on the side of the road, which rarely happens. I'm lost and in need of a mechanic. Is that the same thing with AI?

 

You give the example of the pilot. I get it that a pilot needs to know how to ride a plane, but all the passengers that are in the plane, we don't need to know how a plane works, right? So, will that cause a, all of us, as employers and as workers, to readdress the kind of skills that we really need to know?

 

David Autor [00:11:28 - 00:11:35]

Absolutely. There definitely will be, right. So, right now, to do software development, you need a lot of formal understanding.

 

Gene Marks [00:11:35 - 00:11:35]

Sí.

 

David Autor [00:11:35 - 00:12:12]

In the future, I think there'll be a lot of software you can develop by basically speaking to a machine and telling what you want it to do, and then you won't need it. No one will think it's crazy to develop software without understanding programming language. It's just like no one thinks you need to understand how to drive a manual transmission to drive a modern American car. You don't have to. But I do think what people need to bring is they need to bring judgment, and this is where it gets tricky, because when you use a powerful tool, with great power comes great responsibility.

 

You can do things very wrong very fast. And if you totally rely on the tool that's not a good idea because these tools are, you know, they make mistakes, like really significant mistakes, and it's actually a skill to learn how to interact with that tool. Just like if you collaborate with a colleague on something, right, you sort of know their strengths and weaknesses. And when they say something, you evaluate, well, is that right? Is that wrong? Do I trust you? Do I trust myself? How do we resolve this?

 

Well, AI is a little bit like that. It's opaque, it has opinions, but, you know. So, my colleagues, Nikhil Agrawal and Tibia Sauls, among other authors, did an experiment with AI and radiology, and there are lots of scan reading, x ray reading software now, like the one they studied is called Checkspert, chest expert, and it's pretty good. With just the scan alone, it's as good as six out of ten radiologists. So, you might think, now, I should say radiologists also have a lot of contextual information. It's not just the scan alone, sure. But you might think, well, therefore, if it's pretty good, then doctors and radiologists, your radiologists and software would be even better.

 

But at the moment, radiologists do worse with the software than they do on their own. And the reason is because they don't know how to use it well. So more specifically, the software has uncertainty, and it tells you how confident it is in a scan. Doctors also have uncertainty. These are very correlated. In general, when the software is uncertain, the doctor's uncertain. When the software is confident, the doctor's confidential. Well, what tends to happen is when the software is uncertain, the doctor is uncertain, the doctor defers to the software. When the doctor is confident and the software is confident, the doctor overrides the software.

 

[00:13:49 - 00:14:34]

Turns out neither is a good idea. When the software is uncertain, you probably have more information. You probably ought to go with your gut. When it's very confident and you're very confident and you disagree, you should ask yourself why you think you're right. And so, at the moment, doctors are actually making worse decisions with the help of this tool.

 

Now, that's fixable, but it requires actual training and practice and expertise in how to work with this tool. And doctors in general are not being given a score. When you give your diagnosis, the software doesn't say, ha, wrong. The correct answer was, it's very difficult for doctors to learn that we need to be careful about just throwing sophisticated tools into the mix and assuming we'll get good results.

 

Gene Marks [00:14:34 - 00:15:08]

Do you think we're talking about AI too early? I mean, I'm working on a piece right now for Forbes, and I have to say, like, so I'm, it's, I want to write a piece about AI and manufacturing, okay. And, and David, I have to tell you, you know, every article that I read about AI and manufacturing, it's all about what's coming, you know what I mean? Like, you can do this, it's going to do this, it's going to do that. But nothing in the real world, unless you're talking about the BMWs and the Amazons and the very largest corporate companies out there that have spending millions on of trying their best to develop this stuff, you know? And do you think that we're too premature with this technology?

 

David Autor [00:15:09 - 00:15:18]

You know, I think this is a well-known paradox that in the short run people overestimate the significance of a technology. In the long run, they underestimate it.

 

Gene Marks [00:15:18 - 00:15:19]

Sí.

 

David Autor [00:15:19 - 00:16:07]

And so, yes, at the moment, the impacts are limited. The potential is very high. And there are some places where it's being used at scale all the time. It's actually being used at scale in healthcare, being used at scale in customer service, in financial decision making. But in a lot of other places, people are still just beginning to figure it out. But it will require a lot of experimentation to figure out how to use it well.

 

So, it's appropriate for organizations to be investing in saying, hey, how do we use this? What works for us? In the long run, you're really going to want to use this tool, but don't expect it to work miracles for you. How to even figure out how to integrate it into your workflow is a very difficult problem. I'm struggling this all the time. Like, when should I. Am I using this tool too much? Am I not using enough? How do I use it when it's actually useful to me?

 

Gene Marks [00:16:08 - 00:16:32]

Right. I feel like it's added like an add, like an extra layer of stress into our lives, like we should be doing something with this, you know, that, because we don't want to miss the boat. But remember, I've been telling my clients to, you know, they got to interview their software providers, whether it's Microsoft or Google or their accounting provider, their CRM provider, find out what they're rolling out with AI and figuring out where gives you the most ROI. Lean into that. But it's still pretty early days.

 

David Autor [00:16:32 - 00:17:00]

Very early days. Yeah. We are basically at this point it's just untapped potential. And usually most people think about new technology in terms of what does it automate, what can it do for me that I used to do for myself? But most technologies are at their most valuable when they allow you to do something you couldn't do before. Right, right. So, an airplane is valuable because we couldn't fly without an airplane, not because it automated the way we used to fly.

 

Gene Marks [00:17:00 - 00:17:01]

Right.

 

David Autor [00:17:01 - 00:17:15]

And, you know, electrification, you know, didn't just, you know, replace steam engines in factories. It changed the way factories were organized to have much more precise tools at every workstation. Took a long time to figure that out, but that's where the big gains came from.

 

Gene Marks [00:17:16 - 00:18:17]

So, you're, as a labor economist, you know, when you talk to some people that are. That are the most optimistic about AI, they look at it as something that is going to potentially create trillions, hundreds of trillions of dollars of wealth and value globally. It will do things like fly planes on their own and drive trains and drive cars and take away so much stuff that people are doing right now that will be done so much more efficiently that that will create a lot of wealth and value for a lot of people that the world will share. That's the one side.

 

Then you see the other extreme, where a lot of economists are like, this is going to – Goldman Sachs came out with a study, it was like last summer, it was going to eliminate like 15% to 18% of jobs worldwide, which was a giant number. As a labor economist, I'm kind of curious which direction you lean. Do you think that over the long-term, this will be good for labor markets, or do you think that this will create negative disruption in the labor markets?

 

David Autor [00:18:17 - 00:18:26]

So, first, let me say, this is a highly contentious topic, and people in my field have very different views. So, whatever I say will not be representative of anyone's views, except ...

 

Gene Marks [00:18:26 - 00:18:28]

I just want your opinion.

 

David Autor [00:18:30 - 00:21:28]

It's certainly going to create disruption, no question. It'll strand certain forms of expertise and skills that will no longer have market value because they can be done automatically. I don't think we're going to run out of work at all. In fact, we're in a demographic crunch. And unless we dramatically liberalize our immigration policies, which I think we should do, but I don't think we will do, we're going to be running out of workers before we run out of jobs.

 

So, the question is not whether we'll have jobs, but whether they are expert jobs that pay well, or whether they're generic jobs doing residual tasks that machines can't do. Right? So, a world in which everyone is waiting tables and washing dishes because we haven't figured out how to get robots to do that yet, is not as good a world as one in which everyone is doing healthcare and air traffic control. And the difference is expertise. Human labor is.

 

Everyone can – sorry, not everyone – most people can do table waiting and dishwashing with little training or certification, which means it won't pay well. Expert work is work that creates valuable products or services, and that not everyone knows how to do. And so, we have a major stake in the value of labor expertise and a world in which all the, you know, labor, all the value is coming from machines, right? That's a very difficult world to govern.

 

You have what's called, what economists call resource curse. It's like all the money's coming from one place, a hole in the ground. If it's oil, that's very appropriable. It allows a government or bad actor to essentially get a stranglehold on that and command the resources for him or herself and not really tend to share them, and that's what's happened throughout most of history.

 

So, I believe democratic societies are strongly dependent upon well-functioning labor markets in which labor is valuable. Right now, about 60% of all national income goes to labor. And you might think, oh, in rich countries that have so many machines, so much capital, that would be a small share. It's actually larger. So, the labor share in the US is like 60%. In northern Africa it's like 40%, right? So, as we've advanced, we managed to make human skills more valuable, and that is really foundational to democracy.

 

So, I do not look forward to a future – I don't think we're going to get to one, but I don't want one – in which all the work is done by machines and people are just waiting for their UBI checks. I think the scenario people have in mind is Wall-E, where fat people sitting on armchairs drinking big gulps and watching holographic TV. I think that's the good scenario. It's not even a likely one. The more likely one is Mad Max Fury Road, where you replace the word gasoline with compute and everyone fights to the death over it. But again, so sorry, that's kind of a rant, but let me say ...

 

Gene Marks [00:21:28 - 00:21:28]

That's good. It's good.

 

David Autor [00:21:28 - 00:21:58]

The goal of technology should be to enable new capabilities, not simply to automate. If you were automated all of ancient Greece, it wouldn't be modern America, it would just be ancient Greece without horses. It wouldn't have penicillin, it wouldn't have electrification, it wouldn't have telecommunications, it wouldn't have flight, wouldn't have indoor plumbing. Most of the ways that technology has dramatically improved our lives is by creating new things that we can do.

 

Gene Marks [00:21:59 - 00:22:11]

Agreed. And when you look at all the jobs that existed 100 years ago, 200 years ago, that don't exist now, horse and buggy drivers, blacksmiths, people that tapped on windows to wake people up in the morning before they invented alarm clocks.

 

David Autor [00:22:11 - 00:22:11]

That's right.

 

Gene Marks [00:22:12 - 00:22:26]

People evolved. My other sustaining thing about the humans wanting to be busy is when the tax code was first released, it was, what, two pages? People will find something to do regardless of the technology.

 

David Autor [00:22:27 - 00:23:11]

Well, a lot of the work that we have right now is new work. So, I have a recent paper with colleagues Caroline Chin, Anna Solomons, and Brian Segmuller that shows that about 60% of the work that people do in 2020 didn't exist in 1940. And a lot of that work requires new expertise. Right. We didn't have pediatric oncologists. We didn't have wind turbine technicians. We didn't have remote vehicle pilots. We also didn't have behavioral therapists. We didn't have dance, well, we had some dance instructors, but many services dedicated to the elderly, to the young. And so, a lot of what we do as we evolve is we create new things that require human expertise that are valuable, and that makes for good jobs.

 

Gene Marks [00:23:11 - 00:24:15]

Good. I agree. Okay, let me pivot a little bit, because we've got about 10 minutes left. I had more questions than we're going to have time for, but I got to pick what I want to ask you about. I'll tell you what's been on my mind only because I'm thinking of writing a piece of this for The Hill is minimum wage.

 

And the reason why I bring that up is the debate continues on. I live in Philadelphia. Pennsylvania right now is considering an increase to their minimum wage. This debate goes on and on. Obviously, you know, the federal minimum wage is still at $7.25 an hour, and more than half of the states have increased minimum wage this past year alone. So, it's a continuing debate.

 

You know, and I gotta tell you, Dave, like, I'm not an academic like you. I go and I do my research and, and I find studies that say increase the minimum wage means that employers are going to fire employees or not hire as many people. And then I can give you an equal number of studies that say no, it has no impact at all on employment. It's debate, and I don't know if there's any right or wrong answer, but again, this is a conversation just with you, and based on your research and your thoughts, what are your thoughts on minimum wage?

 

David Autor [00:24:15 - 00:24:26]

Sure. Well, I mean, let's go. There's a level of the minimum wage when all economists agree it would be a bad thing, right. If we said the minimum wage is now $60 an hour, right? That would create mass unemployment.

 

Gene Marks [00:24:26 - 00:24:29]

That's actually ... keep going.

 

David Autor [00:24:29 - 00:25:18]

The minimum wages we've seen in the United States over the last 40 years have been much, much more modest, and the wage increases that we've seen have been relatively small. The vast bulk of the reliable evidence suggests that those wage increases - raise wages - and have very limited effects on employment. And right now, in the current labor market, the current minimum wage is in many cases irrelevant because wages have risen so fast since the pandemic among low-wage workers that most of the wage increases we're seeing are not due to the minimum. And the minimum wage is not buying out.

 

And it's important to emphasize the federal minimum wage being $7.25 an hour, that's in nominal terms, and that was passed during the Obama administration. That $7.25 might have been a meaningful number back in 2012. It's not a meaningful number now.

 

Gene Marks [00:25:18 - 00:25:36]

Wait, if I can interrupt you, though. I mean, you do raise the point that hourly wages, and particularly for blue-collar workers, have increased significantly. Definitely. So, isn't that an argument against minimum wage? Isn't that just like, hey, let the market takes care of itself because that's exactly what it's been doing?

 

David Autor [00:25:36 - 00:26:43]

Well, it's only been doing that the last few years. And in fact, in the pre-pandemic years, almost all the wage increases at the bottom were coming from the minimum wage. So, I think we haven't, I know small-business people will not be happy to hear me say this, but we don't have enough competition in low wage labor markets in the U.S. And so that absence of competition not only allows wages to be too low in some sense, but also allows, gives firms more leeway than they should have, and you have unproductive firms paying low wages. You have more productive firms next door that are paying higher wages. And you would say, oh, how can that even coexist? That can't happen in my Econ textbook. But the world is much more frictional than that.

 

Minimum wages actually tend to weed out less productive firms, reduce market power, and move workers towards the better ones, up to a point, right? You can go too far. You can go overboard on this. We haven't seen much of that in the US because the minimum wage is so modest. It's much more aggressive in many other countries.

 

Gene Marks [00:26:43 - 00:26:50]

Do you think a minimum wage should be a national minimum wage, or do you, and again, your preference, do you think it's more of a local issue?

 

David Autor [00:26:51 - 00:27:30]

I think there should be both. I think there should be a federal minimum wage that sets a floor, and I think it should be indexed to inflation so we don't have this ridiculous, set it, then down it goes. And then I think states, but it should not be as high as the highest state minimum wage. And then states should have discretion to raise it above. The wage structures really differ across states, right. The wage structure in Washington state is very different from the wage structure in Louisiana, and they shouldn't have the same minimum wage. And it would be destructive if Washington's minimum wage laws were imposed in Louisiana. But I think we should a federal minimum wage that at this point it could be $12 or $15 an hour and index to inflation would be a very good floor.

 

Gene Marks [00:27:30 - 00:28:01]

One of the things I hear from clients and people in my community that business owners is that say the federal minimum wage did go up to $12 to $15 an hour, and again, that would impact about half the states in the country, doesn't that have an upward pressure on all wages? You got that person that's making $20 an hour when the minimum wage is $7.25 and now the minimum wage is $14, that person making $20 an hour is going to be like, hey, I should be getting a commensurate increase as well. And that's why a lot of business owners have an issue with it. What are your thoughts on that, do you think?

 

David Autor [00:28:02 - 00:29:09]

Yeah, the minimum wages do spill upward over time. Essentially that, you know, it's like, hey, I used to behave, you know, I've been here two more years. I was making more than the person below me. Now they're making the same as me. Don't I deserve something for that? And so, yeah, I think that's right.

 

So, there are wage norms and expectations having to do with hierarchy and, and you should be paid more than the person that you're supervising. So, I think that's right. And again, you could go too far, but, so, I don't have a, I don't have, there is no correct answer to this question, but the evidence about what the exact right level should be. But the evidence that we've seen says it has not. In the U.S., in the data of the last 40 years, we've really not seen minimum wage episodes that have reduced employment. They have raised wages.

Now, that doesn't mean it's good for employers, right? It cuts into profits often. It also causes employers to raise prices of goods and services, so it gets passed through to customers. You know, in general, the low-wage workers are lower paid than the customers, the businesses they serve. So, it's a form of redistribution. It's not free.

 

Gene Marks [00:29:10 - 00:29:33]

I found that's anecdotally, that is what I found, as well. I don't see many of my clients going out of business because of an increase in minimum wage. I mean, you're running a business. You can do the math. So, they're going to respond in some way and it'll either be cutting overhead or increasing prices or maybe getting new technology to save time. But in the end, it's not going to cut their employment per se. It's going to more have an impact on their profits, like you said.

 

David Autor [00:29:33 - 00:29:39]

Sure. Again, if you said it $60 an hour, it would, but we haven't seen anything that looks like that.

 

Gene Marks [00:29:39 - 00:29:52]

Great example. All right, final question then I'll let you go. I wanted to talk about the workforce today, David. We went through the great resignation after COVID. That seems to have sort of leveled out a bit, right?

 

David Autor [00:29:52 - 00:29:53]

I mean, people are back.

 

Gene Marks [00:29:54 - 00:30:33]

People are back. And not only that, but there is again, to me, I'm not academic, so it's all anecdotal. It's based on my clients and the people that I interview and the people I see. Workers are more concerned about losing their jobs now than they were a couple years ago. They see technology, they see the tech industry and other industries laying people off. I'm seeing less quiet quitters and coffee badging than I saw a year or two or three years ago.

 

Is that, as a labor economist, is that perception in your opinion, correct? Do you think that the markets are tightening now even with such low unemployment, or do you think we're going to have an ongoing labor shortage that we've been experiencing?

 

David Autor [00:30:33 - 00:32:10]

So, it depends where you're looking. So, in this lower end of labor market, people who are non-college educated, doing services, food service, cleaning, entertainment, security, some environment, there's remains wage pressure. Labor markets are tight. Wages have risen a lot. They're not rising more.

 

So, I don't think it's the same kind of rate of change it was back in 2022. But they've stabilized at a much, at a higher level and they've grown much more than, significantly more than inflation has. I think in the white-collar sector, things are not as robustly tight. And we see firms rethinking about do they need college graduates for all this? We certainly, in the tech sector, there's a lot of downsizing. Now, whether that's AI, which people, firms like to claim, "Oh, we're just getting new efficiencies", or it's whether they over-hired during the pandemic that's less clear.

 

I think demographically we're going to be in for a tight labor market, for hands-on service jobs and young people for a long time. To come unless we dramatically change our immigration policies. So, the U.S. labor market is in really good shape relative to other countries. We have very low unemployment, high labor force participation rates, the black-white gap in earnings at lowest level in recorded us data. Productivity growth is strong, and inflation is mostly under control.

 

We have come out of the pandemic so much stronger than the countries to which we compare ourselves. It's really remarkable. We, you know, who gets to claim credit for all that? I don't know, but we're glad to be in that situation.

 

Gene Marks [00:32:11 - 00:32:27]

Dr. David Autor is a Ford Professor of Economics at MIT. He specializes in labor economics, macroeconomics, the political economy, and public economics. David, thank you. Great conversation. More questions to ask you. Maybe we'll get you back in the future, but I really appreciate your time.

 

David Autor [00:32:27 - 00:32:30]

Thank you for inviting me, Gene. It was a pleasure.

 

Gene Marks [00:32:30 - 00:33:22]

Everybody, you've been watching and listening to the Paychex THRIVE podcast again. If you'd like some advice or tips help running your business, sign up our Paychex THRIVE newsletter. It's paychex.com/thrive. Paychex.com/thrive. My name is Gene Marks. Thanks for joining us. We'll be back again next week. Take care.

 

Do you have a topic or a guest that you would like to hear on thrive? Please let us know. Visit payx.me/thrivetopics and send us your ideas or matters of interest.

 

Also, if your business is looking to simplify your HR, payroll, benefits, or insurance services, see how Paychex can help. Visit the resource hub at paychex.com/worx. That's W-O-R-X. Paychex can help manage those complexities while you focus on all the ways you want your business to thrive. I’m your host, Gene Marks. Until next time, take care.

 

Announcer [00:33:24 - 00:33:29]

This podcast is property of Paychex, Incorporated, 2024. All rights reserved.