It’s a perennial story. The demise of mom-and-pop stores, the shuttering of small businesses across the country, the undercutting of local sellers by massive corporations – in short, the retail apocalypse. But economist Michael Mandel has looked at the data and finds a much different story. Not only is the economy adding more jobs than it’s shedding thanks to the massive online retailers allegedly wreaking havoc across the United States, these new jobs actually pay more as well.
His work cuts against the conventional wisdom, and he joined the show to explain his case. We also discuss America’s productivity problem, whether the official stats are mismeasuring it, and how we think we’ll look back on the economy 25 years from now.
Michael Mandel is the chief economic strategist at the Progressive Policy Institute. What follows is a lightly-edited transcript of our conversation. You can download the episode by clicking the link below, and don’t forget to subscribe to my podcast on iTunes or Stitcher. Tell your friends, leave a review.
When we talk about technology and a lot of the changes in the economy – AI, robotics, what’s on the horizon – there’s a lot more concern than anticipation right now. One of the concerns is the job market: the robots are going to take all the jobs. And if people are trying to look for evidence of that, one obvious place to look is local malls, where they see stores closing down or struggling to stay afloat. And in the meantime, they’ll see stories of Amazon becoming a trillion dollar company. People call that the retail apocalypse, the demise of brick-and-mortar retailers and the elevation of online retailers, with all the job loss that ensues. What is your take on the retail apocalypse?
Well, let’s be really clear about this. There is no sign of the retail apocalypse in the data at all. There are two things going on. First, if you look at brick-and-mortar jobs and full-time equivalent jobs, they’re basically flat. They’ve been flat over the last two or three years. They’re basically back to their level where they were in 2007. There’s no great sign of job loss, job erosion, people being laid off. Yes, some chains have closed. But if you ever looked at the Wikipedia page called defunct department stores, it goes on for pages. All of the department stores of our youth are gone. This is the way the world works. Some of the some of the chains that have struggled most like Sears have been going under for a decade.
And so you have growth in some areas and then you have shrinkage in other areas and it basically evens out. One place we’ve had expansion, for example, is truck stores. The big chains have expanded a lot. They sell a lot of food. They sell a lot of groceries. These are places now that people go to pick up items where they used to go to the corner store before. There is a lot of substitution here.
So there’s no sign of erosion in brick-and-mortar. On the other hand, there’s been tremendous growth in e-commerce jobs. And e-commerce jobs are not just in Amazon headquarters, but they’re in fulfillment centers which often employ up to five thousand workers and in many places these are the biggest job projects that have been in the area for a while. And then there’s the drivers that bring the stuff to us. And really what’s gone on here is not a substitution of e-commerce for brick-and-mortar, but there’s been a substitution of e-commerce for our hours.
It used to be that you drive to the mall, take a couple of hours, circle around, and park. Then you wander through the mall and you pull stuff off and you stand in line. Then you drive back home again. Maybe it’s three or four hours altogether. Now what we’re doing is we’re paying people effectively to do the picking and packing and driving for us. So there’s been a mammoth shift of hours from the non-paid household sector to the paid market sector. In some sense e-commerce has turned out to be a machine for generating jobs, not destroying jobs.
And if you think about this in the historical context, this is exactly what happened in the industrial revolution in the early part of the 20th century. The new factories of Henry Ford and GM were highly productive and they were so productive that they could offer cars at a relatively low price that people could buy, at which point the demand accelerated and the number of jobs increased. There is no historical rule that productivity gains cost jobs. And what we’re seeing right now is a living example of that.
I’m saying online retailers when I say e-commerce but is it really just Amazon?
No. A lot of Amazon’s sales are actually not Amazon. They are third-party sellers selling through Amazon’s system, which is not just the website, but in fact their fulfillment centers. And this is really important because these fulfillment centers are not warehouses. They are extremely efficient machines for taking single items and shipping them out to consumers. And so this is a tremendous gain in productivity that the rest of the retail sector is taking advantage of, and the next stage is going to be new manufacturers taking advantage of this as well to ship directly to customers. What Amazon has managed to do is transform not retailing but actually distribution.
The other thing that happens, which is really important, is that you have a rise in wages. Think about what was going on in retailing for the last 20 years. Real wages were flat, and this was an important part of the rising wage gap in the economy because you’d have the industries where wages were rising and then you’d have retailing where the wages were flat and was the job of last resort.
Now what’s happening is that e-commerce is working with more technology. They are better paid. They have better benefits. These are real jobs as opposed to part-time jobs with no benefits.
I have heard these are terrible jobs though. The rap on them is these are just brutal, difficult jobs and they are not “good jobs.”
They are not office jobs. They are mixed physical-cognitive jobs. They require physical work. These are the sort of jobs that we were mourning before in the factories. The steel mill jobs were physical jobs. What we were bemoaning before was the fact that the technological revolution was not creating jobs for people with a high school education. Well now it is and people are complaining they’re not office jobs. Duh. They’re not office jobs. To work in an office job, you need to have a college education.
And so what we have here is we are creating exactly the sort of jobs that we want at a reasonable wage. My data was consistently showing that e-commerce jobs were paying 30 percent more than comparable brick-and-mortar jobs in the same area. Amazon helped by saying they’re going to pay everybody $15 an hour, whereas companies like Walmart were paying $9 an hour. People will say $15 an hour is maybe not that much in Washington, DC, but it’s a darn good wage in Kentucky or Georgia or other parts of the country that have lower wages. And so that addresses geographical dispersion as well.
This is a really interesting time because I actually think what we’re going through here is a way-station. We have the transformation and the upgrading of distribution systems, and the next step will be an upgrading of manufacturing, and that’s going to be really significant. What’s going to happen is that we’re going to see new manufacturing operations grow up that use these new distribution systems to deliver customized or locally-made goods directly to people. We’re going to have a Renaissance in manufacturing.
Elaborate on that a bit. What would that look like?
So you can imagine for example a furniture factory that rather than shipping in furniture from China or even North Carolina makes furniture custom made for your shape and size. So it’s much more comfortable at the same price. Like an athletic fit couch. And you can’t do that now because the production equipment is not flexible enough. We move to robots and 3D printing and all of a sudden the production equipment becomes much more flexible. These factories will require less people than the comparable factories now, but more workers than shipping it in from China.
I try not to say the jobs will return because they are not going to be the same sort of jobs, but there are going to be goods that we wish we had now and cannot afford to make. I could go down a list in food and clothing in terms of business products, and it’s very much analogous to the shift from the point-to-point telephone system to the internet. The point-to-point telephone system worked but was very inflexible. You need to have wires everywhere going from here to there. Now what happens is we break everything down to bits and bytes and it gets reassembled and is much more flexible. There’s many more things that you can do. And that’s what we call the internet of goods. It has a very different flavor to it than the current system does.
With these higher paying jobs in the fulfillment centers, do you think we’re already seeing that in the aggregate wage data for lower-skilled workers?
I’d love to say yes, but I don’t quite see it yet. I think it may be masked a little bit by changes in the minimum wage. I would expect to start seeing it soon. The numbers have grown enough that it should start making a difference. I would expect to see this start helping narrow inequality. But like I said, the bigger impact is going to be when it spreads to manufacturing, which after all is still a really important industry in this country.
You say there has always been this disruption – department stores have come and gone – but certainly it seems like it has sort of accelerated.
I don’t know; there’s no evidence it has accelerated. No evidence in the data.
What is interesting is that companies like Amazon want to open up physical stores. They want to open up physical stores though that are cashless. And that becomes another interesting flash point here, which is that the transformation moves into how do we actually streamline the financial system so that less of the value-add is captured there and more of it is actually captured in the real economy.
All these things are going on at the same time. Our transformation up to now has been in what I call the digital sector, which includes any industry that has basically bits and bytes. So it’s communication and entertainment and professional services and finance. That’s about 20 percent of the economy. It’s not that much. The other 80 percent has been really resistant and partly the reason why we feel like we’ve been struggling across this period is because 80 percent of the economy has not really digitized.
You’re not going to see the productivity growth you would like, and that’s why this shift into retail for me is incredibly significant for two reasons. One is we’re starting to digitize physical industries, and the second is related: we’re finally starting to create good-paying jobs for workers without college education. That is to me the most significant aspect of this, because how can you say we have widespread prosperity until you start seeing that?
What about other physical world sectors? Education and health care, for example.
Let’s start with mining. The shale oil and gas revolution was really technologically-driven. The ability to do imaging and then horizontal drilling very much technology-driven, very much data-driven, very important.
Then we move into health care and you start thinking, what would digitization look like? Well, we had the first step into this with electronic health records, and the electronic health records still don’t have the data in it that we need to actually digitize the health care system. We don’t have the clinical data. So how do you start connecting up your clinical systems to your claims systems? You start to realize that there are institutional barriers there that gradually need to be eaten away and it’s a process; I’ve heard about digital health records for forever.
I am tempted to think that this is something that’s just not going to happen in my lifetime.
Health care is ugly because in fact the number of health care workers has been increasing faster than the size of the population. In any real sense of the word productivity has been falling in health care – falling. It’s been absorbing large numbers of workers from the rest of the economy, raising costs, and I call it a drag on the economy not simply monetarily, but because it’s absorbing all this skilled labor. We have to figure out how to treat people without quite so much skilled labor – whether that’s through the use of robots, automation, more drugs rather than less, I don’t know; there are a lot of possibilities.
I’d like to turn this whole process of thinking about the health care system on its head. How do you engage in disruptive innovation which includes better care with less workers?
Do you think this country is still as willing to accept disruptive innovation?
We had 20 years where things didn’t work well. We had 20 years where disruption in manufacturing and retail – globalization and the big box store – led to outcomes that on the face of it were not good for workers. The manufacturing number of jobs really fell; in retailing the jobs didn’t fall but the wages didn’t rise. And so it did not look like productivity gains were being passed on, and so we have a 20 year history of skepticism to overcome, back into the Clinton years. We have to acknowledge at this point that the tech boom of the 90s was great. And then it was followed by something that was not so great.
The fact is that multi-factor productivity in manufacturing in most major industries has been pretty much flat since the early 90s. And so from my perspective, I understand the skepticism. I understand the geographic skepticism in terms of certain areas in the country being left behind. That just means we who are optimists have to work harder to convince people with real facts and figures and stories that this is benefiting them, that we are into a new era at this point. As the expression goes, the proof is in the pudding.
You co-authored a paper a year or two ago with one of my AEI colleagues, Bret Swanson, called “The Coming Productivity Boom.” When are we going to see that, and why haven’t we seen it yet?
Well, we’ve started to see it in retailing. It doesn’t show up in the numbers because we’re not counting hours right. Remember we talked about how the shift to e-commerce is sucking away your hours and moving them into the paid sector. We don’t count as part of productivity household hours. So what that means is that our fallen household hours producing the consumption of goods is not being counted. And so there’s actually a fairly dramatic undercount just coming from e-commerce.
One of the areas that I’m looking at is food processing. Productivity in food processing has been extremely weak; it has been negative. Part of that is because we’re improving health and safety. As we digitize food processing their ability to ensure health and safety of the food supply will increase but we won’t actually be measuring that correctly. So right now we’ve got measurement problems in the digital sector and not so much the physical sector. But as the physical sector digitizes measurement problems will increase there too.
What I basically say is, look at what people are doing. If you see people changing their behavior, that means they’re responding to innovation and productivity games. I see people changing their behavior in the way that they shop, and this is because what Amazon and other e-commerce companies have done is basically offer something that is superior on every dimension to shopping in a store, because once they figured out they deliver it quickly and offer free returns, why wouldn’t you shop at home?
How much of this problem is just that it takes time for companies to build up the expertise to learn how to use these technologies and make their companies more productive? Does that explain America’s missing productivity growth as much as mismeasurement?
I think that it just turns out to be very difficult to apply information technology to the physical sector. The way that I always describe this is say, think about Iron Man in the Marvel movies: beautiful graphics on screen and it is very hard to do, but it requires an infinitesimal amount of the processing power that is required to generate Iron Man armor. This leap into the physical world requires a lot more data and gets a lot more complicated. We actually have to get down to being able to work directly with the production process. We also need to see the development of the complementary technologies. And I think the main complementary technology that is now coming online is digital distribution.
So I view e-commerce as an enabling technology for the rest of the internet of goods. It is the ability actually to say, I have a box here. I can move this box from point A to point B really cheaply. It’s not expensive anymore; it is really cheap to sort it. And all of a sudden when you say that I can do things in the real world that I couldn’t do before.
Do you think the gains will be significant enough to change the growth trajectory of the American economy? The Fed, CBO, Wall Street – they all say over the long term this is an economy that will grow 2 percent or maybe a bit less. Can these technologies change that into a 2.5 or 3 percent economy?
I ask the question a different way. The question that I ask is, there are things that are too expensive us for us to do now; will they become less expensive in real terms and enable us to do other things? That includes infrastructure, it includes health care, it includes all the things in the back of our minds that make us think, what would we do if we only had more resources? Well, you go back and you think about the invention of the car and the mass marketing of the car, and it meant all of a sudden you could have houses that were two hours out in the country and still work in the city. You couldn’t do that before.
When I talk about this I tell people, make me up a list of things that you would love to be able to do. And then I will tell you that they’re too expensive right now and I’ll tell you what prosperity will mean 20 or 30 years from now. It might mean the ability to feed people locally. It might mean the ability to deal more easily with climate change. You can go down the list.
I’ll give you a really, really simple one which I like: pedestrian bridges. It turns out that pedestrian bridges are super expensive to build, and so you can imagine the application of digitization to construction would enable us to build more pedestrian bridges.
But that is something that should show up in the data at some point. If all this is happening, it shouldn’t be a 1.8 percent economy.
I’ve written on both sides of this. I’ve written that productivity growth is both under-measured and over-measured. What I decided is I look to see how people are feeling. If they’re feeling crummy, then I’m not going to say that productivity is being under-measured significantly.
They’ve been feeling crummy for the last 20 years. Now the question is right now if I go back to an area where an e-commerce fulfillment center has been set up and people have jobs, how do they feel about this? How do they feel about these jobs? How do they feel about what they’re getting? If the price of food starts to fall rather than rise, will that make people feel better?
At the end of the day I look to see how people are feeling about their lives because to me that’s what we care about. People have been feeling things are too expensive. There are a lot of things we wish we had that we don’t have, and we don’t feel like we’ve got a better living standard than our parents did.
Innovation has been restricted to a very small range. What we in the US do is put our R&D into two areas: information technology and biosciences. And what’s happened to life expectancies in the US for middle-aged people? Not great. So where is the biggest failure? Where is the biggest disconnect? Not the information technologies sphere, which has actually impacted precisely the things it should have impacted – the digital industries.
We’ve invested a lot in biosciences and produced incredible gains on the scientific side that have not yet really translated into the big gains on the output side that we were expecting. The other thing is, those other countries like Japan and Europe that were investing in the material sciences haven’t held up their end. And there’s a lot of reasons for that. But the fact is that I would have expected more. Both of those places have much deeper manufacturing bases than the US. They still may produce the sort of innovations that we need, but it’s been a really slow process.
Innovation is not just a one country thing.
You have to look at how it divides up globally. We have been the acknowledged leader in biosciences. We poured a trillion dollars over the last decade into R&D and biosciences. Where’s the beef?
What should we be doing on the policy side, and does either party have it right?
If you want a pro-innovation policy, I think that the main thing I would do at this point is focus on accelerating digital manufacturing. I’m not going to get overly excited about the big picture. I think we’ve kind of reached a stage here where the technology is starting to appear. We have to encourage entrepreneurship and digital manufacturing. That means exposing people to the new technologies, and giving them help however we can in a way that is satisfactory to the political system.
Is there much to be done politically or is it mainly just a private sector thing?
If you want to jumpstart it, there are places that government action can actually have some leverage. The key is to move the tools out there faster. Give people more exposure; expand education and training and set up centers where people can play with the latest robotics or the latest 3D printing technologies, which are amazing. Get some ideas for products and then go out into the market and raise the money for it with or without some government assistance.
You have to think back and say, how did the bicycle shops of the early 20th century morph into the factory? What are we doing to have that sort of thing happen again?
So yeah, I think there’s a role for government. I wouldn’t be heavy-handed about it. I also think there’s a very large role for the private sector. Now if we talk about health care, in that case what I think is going to break the logjam is that eventually there are things that will come out of the investments in biotechnology that will knock people’s socks off enough to move the needle. I feel that we’re really close to that.
Will they say 25 years from now that we were overly pessimistic? “Here’s what they didn’t see coming. Here’s what they couldn’t possibly have seen coming. To them it seems like nothing but stagnation as far as the eye can see but little did they know . . . ”
I have been a proponent forever of growth and volatility. More than creative destruction on a micro level, but volatility on the macro level too. So it’s really easy for me to imagine that there’s tremendous growth coming in areas that we didn’t expect because frankly if I go back to 2006 and I said what was the single biggest technological innovation in terms of economic impact – one was the smartphone, one was fracking, and neither one was forecast in 2006. Nobody wrote about either of those (with the exception of Bret, which he points out).
And so right now if I had to predict I would say that we are close to a tidal wave of change in the biosciences and a tidal wave of change in bringing manufacturing back to this country. And it may be accompanied by a lot of volatility, maybe not generated in the US, maybe generated globally. And so you sort of sit there and you think, what should we as an economy be doing? Is it more important to stimulate the growth, or is it more important to protect against the volatility? This is one of those things that lays right down along the political fault lines. It is very difficult for people to understand and very difficult to figure out because we end up looking backwards rather than forward.
I think both parties might be split right now between people who fear the volatility and people who say bring it on, that is the only way we move forward as a country, as an economy, and as a people.
You asked me to look back from 25 years. At least politically I say wow, so hard to tell.
This article provided by NewsEdge.