Michael Mandel and Bret Swanson on *The Coming Productivity Boom*

Michael Mandel is the chief economist at the Progressive Policy Institute, and Bret Swanson is the president of Entropy Economics and a visiting scholar at the American Enterprise institute. They join Macro Musings to make the case for economic optimism and to talk about their paper, *The Coming Productivity Boom: Transforming the Physical Economy with Information.*

Read the full episode transcript:

Note: While transcripts are lightly edited, they are not rigorously proofed for accuracy. If you notice an error, please reach out to [email protected].

David Beckworth: Gentlemen, welcome to the show.

Michael Mandel: Glad to be here.

Bret Swanson: Thanks, David. Terrific to be here.

Beckworth: Well, tell us how you got into this topic of productivity. I know Michael's been working on this for a while, but tell us how you got specifically into this paper?

Mandel: That's a really good question. I think what happened is that, the Technology CEO Council, asked us to look into this matter. And we realized that we had very similar views on it, very upbeat, and we're able to combine our respective strengths and produce a paper that both had good economics in it and also a lot of great examples.

Swanson: Yeah, I've obviously been following Michael's work on productivity for years, almost two decades now-

Mandel: Oh, don't make me feel old.

Swanson: No. And, like Michael said that, Tech CEO Council, a lot of tech companies that are producing a lot of innovative goods, they were also seeing the economic data, which was showing a pronounced productivity slowdown. So they came to us and asked, can you figure this out? Can you help us figure this out? We see a lot of innovation out there, but the data don't necessarily show it. And so, I focus a lot on actual technologies and actual firms and brought that perspective. And that's also one thing I've also admired about Michael's work is that, he's an economist who actually focuses on what's happening in the real economy as opposed to just the abstractions of the aggregates.

Beckworth: That speaks to me right there. As a macroeconomist I'm guilty as charged. It's nice to have you guys on the show because at heart, I believe I'm an eternal optimist. But, I've had a lot of guests on the show who have talked about the great stagnation, the slowdown in productivity growth. We want to get into that. So it's refreshing to have someone come in who has a thoughtful piece, that takes a more positive outlook on the future.

Beckworth: And we'll have this article up on our website where the podcast is hosted so listeners can take a look at it. Let me begin by going over the general narrative that's out there, that there has been a slowdown. Productivity has slowed down, trend and growth has slowed down. Robert Gordon has a huge book on this, John Fernald, from the San Francisco Fed has written about this from the perspective of productivity growth. What's going on? What is your take on what has actually happened? Are they right?

The Productivity Growth Slowdown

Mandel: Well, I think you're only giving a partial explanation of what's being talked about. There's two dis utopian visions out there. One is the one that you described, which is no technological innovation, stagnation, slow growth as far as the eye can see. The other just utopian vision, of course, is this, robots are taking over. Boosting productivity tremendously, and killing all the jobs. And so these lie are being said at exactly the same time. And they're two very different views of the world.

Mandel: And, because of the second, robots are taking all the jobs, actually, it's incredible amounts of innovation, incredible amounts of transformation. And it's happening so fast that there's nothing we can do about it. And, what we realize is that, it makes sense to think about the economy as being divided into a digital sector and the physical sector. The digital sector is about 25% of the private sector, and it includes a lot of digital industries where the output is easy to digitize. Not just the internet but financial services and professional services and so forth.

Mandel: And these areas, it is true innovation is happening really fast. And productivity growth is very rapid. Then you've got the physical industries, which includes much of manufacturing, transportation, healthcare, even education, where the rate of transformation is very slow. Where it's hard to digitize things. It's a process, it can be done, but you have to change the way you work as well. And in that part of the economy, productivity growth is incredibly slow. Wage growth is incredibly slow.

Mandel: And so, your attitude towards what's happening, it's like the elephant, right? Whichever part you grasp on to, is what you say is happening. So if you're in the digital sector, it feels like wow, everything is being transformed tomorrow. And then if you're in the physical sector, the rate of productivity growth is really slow and going you look at individual industries and you say, I don't quite understand what's going on here, but doesn't seem to be a lot of change.

Mandel: For example, one of the examples that I really like is, Asphalt, asphalt is manufactured, but price has soared, because you haven't had a lot of productivity gains. And that's why actually a lot of towns and cities have a lot of their roads go back to gravel, because asphalt is so expensive. And you go look around, you can find just lots of examples.

Swanson: Yeah. I think a lot of that, that's very well said. But this also helps explain, I think, some of our recent political challenges. Our political environment and what we saw in last year's election, where there was a lot of frustration across the country, I think, probably due to low productivity and low income growth. But it was there was also a lot of misunderstanding, between different parts of the country in different industries. Between those who, we're not seeing rapid income growth on the one hand and those who are seeing very rapid income growth and productivity growth, on the other hand. And so this may help, I'm certainly no political scientists, but this also may help explain some of our recent politics.

Beckworth: I think it's interesting because those two narratives are inconsistent with each other, right? In the sense that if robots are taking our jobs, why don't we see that in the aggregate numbers?

Mandel: That's right. And it turns out if you actually look at the amount of investment in robots, it's really low. And most of is going into the motor vehicle industry.

Swanson: Yeah, no, I just think the robot narrative is probably in the opted pages, has probably outstripped the reality.

Mandel: I think it turns out to be harder to transform physical industries, and you think with digitization, because the physical world is very complicated, think about autonomous vehicles. We can put an autonomous vehicle out on the street, can it react to every possible instance? What happens if it gets broken down? How reliable does it have to be? And, if you think about software programs getting flaky? Well, you can't have a situation where your car gets flaky on you. You're going to have this vehicle, which is this combination of a complicated physical system and complicated software system.

Mandel: And yeah, we can eventually get it right. It's going to take a lot longer than people think. And I think that our biggest challenge economically, is how do you take physical industries and digitize them? And a lot of what we wrote about in the paper was examples of how that's happening, which has the effect of boosting productivity and often creating new jobs.

Beckworth: I want to go back to the apparent slowdown in the aggregate numbers. So this robot story, the robot narrative, the people who are in the digital industry, they see it. But would this suggests that's a small part of the overall economy? Or is there more in the overall economy, but we're just not measuring it? Because there's been this debate that maybe there is more rapid gains. We're just not seeing the actual measurement because of problems in GDP?

Swanson: I think there are some very real measurement issues. Lots of really good economists have looked into this. I think we probably do underestimate innovation and productivity in some sectors. It's hard to count how health advances. Some of those don't show up in the data. There's very good work showing that we explicitly miscount, overestimate the prices of many semiconductors. And yet the same economist who did that work who showed this mismatch measurement, also looked at how that fit into the overall economy. And they say even if we are under estimating productivity in these various ways, and we give productivity a big boost, we still see it pronounced productivity slowdown over the last 10 years.

Swanson: So, even if you account for those measurement challenges which are very real in some cases, that means we would have had to have been counting correctly in previous eras. But there probably was mismeasurement back then, too. So the bottom line is that we think there has been a productivity slowdown.

Mandel: Let me just add on to that, I could regale you for hours with examples on both sides, under measurement and over measurement of substantial size. And some of them are really prosaic, the price of burial caskets has gone way up, is that reflective of real price gainers or has there been a quality improvement? That's a really hard question to answer. But, at the end of the day, I look at the reaction of people and how they feel. And it's clear that people feel like their real living standard has not gone up, and they're worried about their kids. So when I sit down, and I ask this question, I say, in the end, the slow down looks real because that's what people are telling me.

Beckworth: All right, so at least it's perceived as real. And these studies that you've mentioned, Bret also suggest that it's real as well, even after you account for mismeasurements.

Mandel: Yeah, I think that there's deep problems in our statistical system at this point in terms of how do you deal with a data driven economy, both the plus and the minus side? And, then also how you think about life expectancy, because in fact, we know about the fallen life expectancy for certain groups. And honestly, we should be counting that in as well. And so at this point now, well really, if you thinking about this mismeasurement, you don't really want to think about mismeasurement anymore. You want to think about, what new framework do you actually need to deal with a data driven economy where your productive activity is not simply located in one country? We do the best we can, we use at this point the conventional statistics like everybody else does.

Mandel: But you got to keep in the back of your mind that, this is not mismeausrement in the conventional sense, but it may be that we actually start need to thinking about a new framework for a data driven economy. What are we going to measure? What are our indicators going to be may be very different than they were in the past.

Beckworth: Now, for our listeners, I'll let them know that Michael has been very vocal about this issue in the past, I remember your earlier work on measuring the productivity boom, early to mid-2000s. So you've been pushing for better measurements. So where would you have them go? Do you have any suggestions in terms of a better measure of welfare or progress?

Constructing Better Measures of Productivity

Mandel: Well, let me just make the first obligatory statement, there's very smart statisticians in the government. And if we give them more money, there's a lot that they can do to improve what we have. The real question here is what is our goal? Because GDP, if you remember, was invented in the 30s and 40s, for two purposes. First, to track the downturn, the depression. And second to assess the war fighting capacity of the US and the 1940s. And if you think about the second one, it's interesting, what would we need to know now to understand what our capabilities are, because we'd have to think about what kind of war we want to fight.

Mandel: And then the third question that people would get is standards of living. And you can go back and forth, but I suspect Bret would... and you to think about things like life expectancy, you have to think about, education. There's a lot of things you want to think about as part of a good life at this point, which is not in this statistics right now.

Beckworth: Well, something I've wondered, and I've asked some previous guests, I had Jason Furman on I asked him because he's with a CEA. I asked you guys, since you guys are technology specialist. Why don't we have more real time measures? So I have friends who work with hedge funds, and they look at real time debit card, credit card, consolidated. They can tell instantly what's happening. And there's big seasonality issues. You're not maybe getting truly value added measures, you just getting transactions. But why isn't there some weekly, some higher frequency measures that we're not getting from the government right now?

Swanson: Well, let me just give you one example. It's a really good question, and I think we could get more data if we try, or maybe it is an issue of money setting up new pilot programs in some of our economic bureaus and so forth. But one of the studies I referred to, from David Byrne and Steve Elinor and Dan sigil.

Swanson: They found a specific problem with the way we were accounting the prices of microprocessors over the last 15 years. Where the official producer price index says that, there is an annual 6% decline in the price of microprocessors quality constant. Where, if we believe in Moore's law that should have been much larger. Well, what happened about in about 2005 Intel, the world's biggest producer of microprocessors changed the way it basically priced and the government kept on using the least prices. Even though the actual prices that companies building PCs or building servers in the cloud paid for these microprocessors changed dramatically compared to the least price.

Swanson: So when they went out and found used actual prices over the last, say dozen years, they found that microprocessors had been declining 42% per year.

Beckworth: Wow.

Swanson: So you compare 6% to 42%, that basically tells you that Moore's law in effect is alive and well. One example of exactly what you're saying using more actual, not real time, but closer to real time data.

Mandel: We can get real time data now. Real time data is great. But the time you really want it is at a turning point where things are changing very rapidly in the economy. And unfortunately, that's also the time when relationships between the real time data and the conventional macroeconomic aggregates breakdown as well.

Mandel: So, in my view, actually, the ideal situation is times of normalcy, to rely on a lot of the real time data, and then to use the government surveys just to benchmark. Then actually what you need is more government surveys, when things look like they're changing very fast, because that's how you can really track it. Because they're real time relationship, just like the financial markets broke down. The real time relationships and the data breakdown too.

Mandel: So you got to be clever about the way you integrate the real time data, what they call administrative data or organic data, with the government survey data. The government survey, you actually know what you're looking at, because you've got the right sample frame. With the administrative or organic data, you don't know exactly what you're looking at. It's cheaper to use. It's more real time, but it will break down unpredictably.

Mandel: And, I should say, at this point, I've got a grant from the Sloan Foundation to have a look at actually, at some of the uses of organic data, and just this just way. Fascinating subject, I could go on forever.

Beckworth: Well, that's unfortunate that it does break down at turning points, because that's when you need it, because, the surveys are going to take time, right?

Mandel: Not necessarily. You could put in a capability, in times like 2008, for doing multiple fast turnaround government surveys. You could do that. If you were set up, this I mean, we have to be clever about this and figure out what we're really concerned about. Because going back and looking you'd say, that's when we needed better data. We can ask a very specific questions, not an abstract question, what could we have done over that period to give us a really close measure of what was going on?

Mandel: And I think the simplest idea then, is actually supplementary government surveys, because my sense is that, the administrative organic data becomes less reliable. What happens, for example, in job postings is that, you may have companies posting as many job postings that they did before, but also they're not filling jobs. And you won't know that until afterwards. So, traditional relationships will break down or in the credit card aggregates. Well, when you see this sharp slowdown, is this actually assigned to the economy is collapsing or is this a sign of good adjustment?

Beckworth: Right, and there's structural change too, there are people using more credit cards, using cryptocurrency, there's a lot of things that happen that we can't-

Mandel: Quickly at turning points. And so, this is what I mean is when you think about redoing our statistics, the needs of real time macro policy are very different than the needs of defense capabilities. And are very different than social indicators. And the three shouldn't be confused with each other.

Beckworth: Well, Michael, we look forward to your report on this, pushes in the right direction. Well, let's get back to this big point that really motivates the study, at least in part and that is the slow down. And what I've heard you guys say so far is it is real, even after accounting for measurement problems. And I had a hard time accepting this at first, I'll tell you my personal take. I look at my smartphone, all these apps on there that do things that were formerly a part of GDP, everything from music, to books, to compass, to map, things we used to pay. So I bought, yeah, it's understating, but these articles, these better comments put out to convince me that there's more to it than that.

Beckworth: I think what really hit me though is two things. I mentioned this in the teller counter interview number one, if you look at total factor productivity, trend growth relative to where it was going before, the 70s, it's clearly has fallen down. So the question is, not that yeah, we live in a great time of improvement. It's where we could have been had we kept up that pace, I think. And I think of technology. So Peter Thiel has this famous saying now, famous quote, we wanted flying cars instead of we got 140 characters. Why aren't we in that alternative universe? Why didn't productivity growth keep up and deliver us that reality versus the one we have?

Could Productivity Growth Have Been More Robust?

Swanson: I think that's a great question that gets to the heart of our study, and, it's the atoms versus bits perspective that Thiel has put forward. And I think that encapsulates a lot of what's going on here. So part of its what Michael was saying before about, we had these new digital tools with the computer revolution and the internet. And the easiest places in which to use those tools were the already information rich parts of the economy. So whether that was media, whether that was finance, things that were easily digitized, as he said. That the physical world just turns out to be much more difficult to exploit with these new information tools. That might be part of it.

Swanson: I think another big part of it, is the fact that a lot of these physical industries are also much more heavily regulated. So it's much more difficult to run new experiments via entrepreneurial firms, launching new products and services, in some of these fields like health care and education and manufacturing and energy. So it turned the digital world the digital economy also happened to be part of the economy that had the most free rein to experiment and innovate.

Mandel: I would add at this point is that, the process of digitizing physical industries, well, you mentioned your smartphone. Your smartphone is great for looking at videos and downloading information. You can't actually teleport food onto it. It just doesn't work that way. If you order a food on it, it still has to be delivered by car. All right, it has to be made in a kitchen. And this was the mistake actually that the early eTailors made in the 1990s. They said okay, we'll slap up a website and we're done. This we've now digitized retail.

Mandel: Well, it turned out that, retail was a little bit about how people order and a lot about how the goods get delivered, get from the manufacturer to the person. And the initial retailers, just died for the most part, with Amazon being a notable exception.

Mandel: And so now we are 20 years later. And what's happening is, we're actually finally seeing the digitization of retail. And it turns out not to be about digital per se, but turns out to be building these high tech fulfillment centers, which can turn around and get goods to people in a day. The turning point of this was in realizing, when people shop they want immediate gratification. And immediate gratification is defined as very rapid, and that's very difficult to do, and it requires really sophisticated digital engineering, and it also employs a lot of people. And it raises productivity.

Mandel: And incidentally, it may create new capabilities that the manufacturers can use, to offer custom manufacturing to people with really fast turnaround. And so we're looking at this point is actually the digitization of retail, which has the effect of creating jobs, creating better paid jobs. But it looks different than what we expected it to look. It involves a lot of physical plants being built in places like Tennessee and Kentucky and Indiana, and Ohio, and Oklahoma and Pennsylvania. It involves physical-ness. And that's what I think, Bret mentioned this, the essence of this paper is saying, turning atoms into bits, it's a process. It's a long process but it's happening. And as it happens, it has the effect of raising productivity. And incidentally, creating jobs as well.

Beckworth: You mentioned Tennessee, I live in Tennessee's I think my listeners know. And there's an Amazon warehouse not too far from me in Lebanon, I live in Nashville, Lebanon's just outside is huge, huge. And so that facility is built. I'm just waiting for the day when the drones will pick up a package from that warehouse, come to my home and drop them off.

Mandel: Well... what you really want to think about, is you go back to the development of the auto industry. There weren't actually all that many people who worked in the auto industry. What was important was all of the spin off industries that were created. So, right now we're building an advanced distribution industry in the US, which is able to get stuff to people really quickly.

Mandel: Well, does that mean that all the center manufacturers want to locate near the advanced distribution facilities, so they can quickly turnaround stuff to consumers? And will that create a new and different type of manufacturing in the US, that has a durable competitive advantage against foreign suppliers? I mean, you have to step back and say when we transform physical industries, and I know that Bret has a lot to say about this. Bret supplied all sorts of great examples for this paper, about what you can actually do, not just simply substitution of capital for labor, substitution of robots for people. But we really look historically at productivity growth, you create new capabilities that get used in ways that create new industries and new products, new services that employ more people.

Beckworth: Well, let's move into that, in your paper, you list some key technology areas that are driving change, and I want to read them off and then you tell me in a practical sense, how these things are actually being developed and seen in the physical world. So you mentioned mobile technology, big data, ubiquitous sensors, computer vision, robotics, artificial intelligence, 3D printing, 5G wireless. So you draw upon some of those Bret, tell us how we are beginning to see some of the transformation in the physical world through them?

The Technological Developments Driving Change

Swanson: Sure, just a few examples from the last few years, where we're getting some really promising signs of what is coming. I give two examples, one in transportation, so the rise of Uber and Lyft and the first data is ride sharing services, that have just exploded. And the first data we're getting from them, actually shows that they are far more efficient than the old taxi services. So they actually carry people in the cars, a far greater percentage of the time that they are running. Some, 40% to 50% more time with a passenger in the car.

Swanson: This has just been, in addition to the ease of use and the convenience and feedback that the digital app gives you. And so this wasn't a huge new technological invention. It was just a smart application of a lot of these mobile and software tools that were starting to be used, and a creative use of a new app, connecting software and mobile with a very physical industry of moving people around cities.

Swanson: And we're already seeing just the tip of the iceberg that it's already more productive. Now you can go on and think about where that's going with autonomous cars and so forth. Another example, where that's very encouraging, is another of the most physical industries, energy extraction. So, in the, mid part of the last decade 2005, 2006 we were supposedly at peak oil. Oil prices were sky high. We were supposedly running out of it. And then the shale revolution hit.

Swanson: And a big part of the... we'd known about shale rock for the last hundred years. That it was a very petroleum rich source, but we didn't know how to get at it. And so through the combination of some very physical processes, so horizontal drilling and hydraulic fracturing, along with very digital processes, namely 3D computer modeling of under surface shale formations, we were able to pinpoint these drills, miles and miles under the ground, and fracture them in just the right place to release unimaginable amounts of oil and gas.

Swanson: And so, in the last decade, we've gone into... the US has basically doubled its daily production of oil, a huge increase of natural gas which is, good for electricity generation. And good for every other industry, we've lowered our CO2 emissions because natural gas has much lower CO2 content than coal. And so those are just two ways that information technologies are transforming some of the most physical industries.

Mandel: Let me add one thing to what Bret said. So the application of information technology to energy extraction, had the effect of increasing employment. Not just of highly skilled Petroleum Engineers, but also the people doing the actual drilling. And this is actually what we find pretty much consistently, forget the fear of the robot people, what happens when you apply information on net, the digital sector is actually creating jobs at a faster rate than the physical sector.

Mandel: Because, the prices are lower, the demand is higher. It's stuff that people want, so we have applied digitization to energy extraction. And we didn't automate away these people, we created more demand for them. I think this is a pretty consistent message. Now you go back and you have the Uber example, we don't have actually the data, that allows us to say has this increased or decreased employment? There's a lot of debate about this, the data is very flaky.

Mandel: And I think what you will find is that if, this is really a truly an increase in productivity, so you can lower the price of rides, it'll increase the demand, people will stop using cars so much, they'll start shifting to whatever the ride sharing application is out there. Whether it's Uber or Lyft or something else. And the demand will go up, because you'll have a transfer of time that people drive themselves, to drivers that are paying.

Mandel: And what that ends up doing is that it ends up increasing the amount of time that people have, and reduces the amount they're going to be paying for transportation, which is enormous share of people's budget, you have no idea. I can't remember the exact number, but big chunk of people's budget goes for maintaining cars, basically. And, if we can take out even a bit of that, you're well on the way of increasing living standards.

Mandel: So, when I look at Uber and Lyft, I see this as the growing pains of the applications of information technology to a physical industry. You started off by asking about flying cars. Well, this is not flying cars, but this is cars that get you where you're going, without you actually having to spend the time maintaining it.

Beckworth: And in Pittsburgh, there's driverless Uber cars, right?

Mandel: Here's the-

Swanson: Yeah, all sorts of pilot program, Pittsburgh, some other places. And there's a useful debate about, in a lot of ways the driverless car technology has come on faster than a lot of people thought. And then there are others that are saying, wait a minute, there's still a lot of issues and challenges that we have to solve to make these things really, truly safe and so forth. But we've also seen driverless trucks, there was the case of the driverless truck that delivered a cold beer across a couple hundred miles on highways and did just fine.

Swanson: So there's a lot of really encouraging evidence on the technology, but it's going to be another, number of years before it's come in place.

Mandel: Let me talk about driverless trucks for a second from both angles, which is really... So trucking has been an industry, which has had a slowdown in productivity growth. And actually the cost of distribution is a big part of that. When you go in and pay for clothing, or food, or anything, the cost of distribution is a big chunk of what you're paying for. The extent to which you can lower that, you can lower the cost to the consumer. If you lower the cost to the consumer, you perhaps increase the demand, you open up the possibility for new markets that didn't exist. I think that's is what we have to think about.

Mandel: When I hear that people talking about, 50% of jobs are potentially affected by automation. I go back to the late 1800s. And you could have written an article saying, 75% of jobs are potentially affected by electrification. Yes, right? It's true. There's no way, that's absolutely true. I could have written, I'm sure people did or by the steam engine. And when you have new technologies come in, they're real quick... I want to make a careful distinction here. There's the macro positive effect. And then what we've learned is we seriously have to worry about people being left behind.

Mandel: Let me actually just before we move on, let me just talk about the extraction. So the number of mining jobs in the country, since 2000, has gone up a lot. The number of coal mining jobs has gone down. Coal mining jobs are concentrated in a couple of states, have really hurt those states. So it doesn't actually help those people. That there's been overall growth in mining jobs. Their communities have been hurt. I think we have to think about that we can have good macroeconomic effects, from productivity gains. Not just in terms of productivity, but in terms of job growth and wages and all the things we love and realize that there are people that are going to be left behind by this. And we have to for a lot of reasons, just take this very seriously.

Mandel: So, we start off talking about optimism. Yeah, it's optimism, but it we understanding that, when people can complain that there are people being left behind, absolutely true.

Beckworth: I mean, a good example, that would be truck drivers. It's one of the few professions left where you don't have to have a lot of education, you can make a decent income. And if it does go the way of driverless trucks. Bret, you said it maybe a few years before we get there. But when this does happen, that's one more job that's disappeared from the lower socio economic employee. Right?

Swanson: It's a real question as Michael was saying, I think you've got to think about though, the other new business models that this opens up. If I can operate a fleet of trucks, maybe I can buy two or three or five trucks, and operate a fleet that is serving a specialized market serving some specific customers. That might be a tremendous opportunity for a former truck driver.

Swanson: So you have to weigh those possible gains against some of the losses. But as Michael said, you may have to think about specific people that are hurt by very rapid dislocations, like the coal miners. But then you also have to think about the overall growth affects that not only will benefit the whole economy, but that will allow us to perhaps offer targeted benefits or transition assistance and so forth to some of these very specific communities.

Swanson: And I think one of the big problems over the last 10 years has been the fact that we've had a few of these dislocations, whether it was from China or whether it's from technology. At the same time, we've had, very slow economic growth, and so they've been a double whammy. If we'd had better economic growth with the dislocations, it would have allowed more job opportunities for people to switch.

Swanson: It would have allowed the, government budgets to help in other ways that government can help. But that's the thing is, that you want a dynamic economy that is growing fast enough so that the inevitable dislocations whether from trade or technology or consumer tastes, we allow for the people most vulnerable to move on.

Beckworth: No, I agree. I didn't mean to make the impression those poor truck drivers that, I believe they to ultimately be better off as this happens.

Mandel: I think you raise a good question. I think what Bret and I both believe, that it's important to seriously consider this. We don't shove this under the rug, we just think that, the history suggests that, the innovations that generate growth, also generate new opportunities. And that create a big enough pie that you can something that-

Beckworth: Well, you mentioned before the show a great example I hadn't heard of. Tell us about Facebook's new hiring, who are they hiring?

Mandel: Well, I think Facebook has been accused of not having enough workers. And, I've looked at this. And basically, if you compare Facebook to GM at the same age, they basically are in the face on the GM path. They've got very rapid growth. They're just relatively young. And now what they've said is that they're going to hire workers to help stamp out fake news. And I think you understand the whole process, you have a new technology, it generates new problems and new opportunities and it generates new jobs.

Mandel: And I think partly what we've said here-

Beckworth: Fascinating.

Mandel: ... I think about this as building guardrails on highways. That if you can go ahead and you can build a highway out there, and you build without guardrails, because why does it even need a guardrail, we're all good drivers? And a couple of people go off the side, and you say, I really need to put out guardrails, you send another crew out to build guardrails. And then you send a crew out to fill in potholes. And every year because you get... and all of a sudden, what looked a one shot becomes a process, becomes an industry, becomes a sector.

Mandel: And you've got the guardrails, and all of a sudden, your aging parents can drive along the road now, and they can go shop wherever they want and they create new jobs. I mean, and the thing is that, we see this process happening. And we see it starting to permeate the physical industries and part of our optimism is, yeah, the process is slow, but it's happening and it will create faster productivity growth and it may very well if I look at, five or 10 years, one of the things that we've learned is that, it's really tough to predict productivity growth 10 years from now from what it was over the last 10 years.

Mandel: And, what Bret and I both believe is that, we're in a situation where the future is happening. And it's not happening in the dis utopian way of stagnation. And it's not happening in dis utopian way of, oh, my God, the robots are eating us all. But in fact, it's happening in a way that's going to raise wages and create jobs that are good overall for people.

Beckworth: Yeah, I just liked that Facebook example, because it's a clear example of new technology creates a new problem, then it creates new jobs to fix that problem. And your point is, this will be a process that continues, as we see more and more of this information applied to the digital world. Bret, you had something?

Swanson: That's right. And yeah, David, I just wanted to step back and go to one of the core findings of our paper, which I think quantifies the magnitude of the opportunity here. So the way we divided up the economy into basically, two groups, the digital industries and the physical industries. The digital industries are about 30% of private GDP output. The physical industries are 70%. But the digital industries make 70% of all the investments in information technology, the 70% of the economy that's physical, makes only 30% with the investments in information technology.

Swanson: And there's a massive productivity differential. So what you are seeing here is basically this is a big problem. It is drugged down growth for the last decade or more, but it's also on the flip side, a huge opportunity. It means that, three quarters of the economy is underperforming and if we could get three quarters of the economy growing faster, innovating faster, that it has large macroeconomic effects.

Beckworth: That's very encouraging. That's low hanging fruit there. I mean, maybe it's not low hanging, it's fruit that needs to be pulled and picked.

Mandel: I think the phrase low hanging fruit is difficult. I don't think what we want to think about, that you build ladders and you climb up and, you've got to take the fruit, and then you got to process it a bit, and so forth and maybe turn it into wine. It's just, in this we're not talking about computer games, we're talking about objects in the physical world. We're talking about people that you got to be careful with. And it's a process that takes time. But, every place we look, we see examples of it happening.

Mandel: And I just wanted to point out one thing, you go back to 2006, Which is just before the introduction of the iPhone. And just before fracking really took off. And you see no evidence in the business press, that these two enormous technological events were about that, actually I should... Bret actually predicted it at the time. Okay, I shouldn't say that, Bret actually predicted it, as he pointed out to me.

Mandel: But it's vanishingly small, our ability to predict technological breakthroughs, even at very short distances is difficult. It's hard to remember just how young Facebook is. It's hard to remember just how young Uber is. It's hard to remember that the smartphone is less than 10 years old.

Beckworth: It's pervasive now but it's-

Mandel: You think about how fast it spread and how many jobs are created for app developers, and so it's... I want to add, we also quantified the amount of growth that we would get over the next 15 years, as the physical sector adopted some of these technologies. And the amount of additional growth and additional GDP and higher wages and more taxes you get is potentially enormous.

Beckworth: Well, if 70% of GDP right now isn't fully utilizing this, that's a huge opportunity, right? Let's speculate here. I want to have a little fun with you guys. So some of these technologies you've listed, robotics, artificial intelligence, 3D printing. Let's just think forward. What will our world look like when 3D printing for example, really blossoms? Will we see declining retail, will we all have 3D machines in our home making everything from tools to clothes? What is your vision of what 3D printing will do for us, Bret?

The Future of 3D Printing, 5G Wireless, and More

Swanson: Yeah, I think you're going to already see lots of hobbyists using it. And I think there will be new businesses started with, creating all sorts of customized products. Whether they're mechanical products, or maybe it's customized clothing, or customized accessories, or customized sports equipment, or customizing parts for your drone. But there are all sorts of, industrial size companies that are rapidly deploying 3D printing, not just for prototyping. Which is one obvious case, it's already being used, but for creating objects that we've never been able to make before, because we weren't able to fabricate them using our traditional casting processes or, forging processes.

Swanson: And that allows, mechanical engineers and others to design machines in ways that were never available before. So there'll be lots of innovation there. I know that some of our local dentists and orthodontists are already using 3D printers to print out appliances based on 3D models of your mouth and teeth in their offices. Which speeds up the process of, and lowers the cost of, dentistry and orthodonture.

Swanson: So, I think that it's going to, like the PC was, there's going to be lots of hobbing that turns into, small businesses and then there's also going to be lots of industrial scale type stuff as well.

Beckworth: I wrote an article about a neurosurgeon who does 3D printing before he goes into surgery, he does a dress rehearsal on. He scans the brain, he prints the person's brain, he works on it. And then he goes in and does the actual surgery. So his productivity shot way up because of the 3D printing.

Mandel: As long as we're in the healthcare, I think one of the interesting things to think about is artificial organs. And as a new manufacturing industry, whether or not they're 3D printed or grown, I'm not sure that-

Beckworth: Fascinating.

Mandel: ... everything has to be, you grow cells on a scaffolding maybe the scaffolding is 3D printed and then the cells are grown. And to a degree to which that this becomes practical. You have to think about the demand. Think about, the creation of a new industry that did not exist before with a new product. This is what Bret was talking about. You can't simply stop and say, oh, we can do what we were doing before cheaper, because that's like saying, okay, we're going to put together an assembly line to build horse drawn carriages.

Mandel: We could do it cheaper, but that's not the point. The point is that we could do something, that we couldn't do before make it cheap enough that people could buy it. And this is the miracle of the smartphone. It's not that we couldn't construct a small computer before, which is what it really is, is that we could do it cheap enough to make it accessible, not just in the US and not just in Japan and not just in Europe, but around the world. To the degree to which this has become almost ubiquitous.

Mandel: And this is really what you got to think about. You have to think about what the future of falling prices look like for new things that we did not know that we needed. That at the point where it comes out, becomes immediately obvious that you need it. And, like I said, there's a difficulty doing technological forecasting, but, nobody knew that they needed Uber, before Uber started.

Beckworth: And what this speaks to is, we need a sense of humility, to not get overly pessimistic about the future, to understand these things are happening. To have a broad vision and a sense of humility that radical transformation can occur. Let me do one more of these industries with you Bret as an example, what about 5G Wireless? How is that going to change our world?

Swanson: Well, in many ways 5G Wireless, the next generation of wireless technology, will be a foundation for many of these technologies and industries. Obviously, the communications network is already a key platform for the economy, but it will only become more so going forward. So whether you're talking about connecting 10s of billions of little sensors in what we call the Internet of Things. As we put little sensors, on trash cans or on pill bottles or on packages or in machinery, or in our bodies. Collecting all this data wirelessly, will require a network that is far more ubiquitous, with more capability that's faster than ever before.

Swanson: And so this new 5G network will be made of millions of new small cells. So you think about cell towers now, in the radius that we now have a cell tower. We may have many dozens of these small cells on street corners, on office buildings, in neighborhoods on campuses, on land poles, utility poles, and it'll be necessary for the Internet of Things, it'll be necessary for connected cars.

Swanson: It will be, in addition to the obvious uses of connecting our phones, and our residential and enterprise broadband. Making sure that we have enough wireless capacity to do virtual reality and augmented reality and, 8K cinema. So it really will be the foundation of so many of these different industries. And it's a project that will take, a couple hundred billion dollars of new investment to build this out across the nation.

Beckworth: I've pretty discussed in the context of driverless cars in particular, that it will be key to making them operational, to connect with each other. But let me ask one other thing. Michael was talking about health, this is not meant to be dystopian, but what about like cyborgs? As humans, we have more computer implants, maybe it's something to enhance my IQ, things like that. Do you see that on the horizon?

Mandel: How would you feel about that? Suppose that I could offer you a pill that would improve your IQ by 25%? Would you take it?

Beckworth: You're very tempted.

Mandel: You're very tempted. And I think we have to understand that we've spent a trillion dollars in the US, on private and public R&D into bio sciences. We haven't seen but a fraction of the payoff of that yet. And as the payoff of that rolls out, there are things that will happen that will raise issues, both positive and negative, that we're not even cognizant of right now.

Mandel: So, productivity is one thing, but remember, with the car, came the suburbs, and came all sorts of other issues. It was a social transformation as well. So, we talking about productivity, and honestly, I mean, I think if you're to ask both of us, we would agree that, the odds of productivity gains accelerating is really high at this point. Then we have to think about what comes along with that, too. It isn't simply about doing what we're doing now, more efficiently. It's about doing different things in the healthcare system look different, the education system will look different, the driving system will look different, retailing will look different. And our goal is to navigate this in a way that improves people's welfare.

Beckworth: Very interesting, Bret, anything else you want to add to that?

Swanson: Well, I just say there are some, as Michael said, and as you suggested, some problems with technological forecasting and what that will do to society and culture and there may be some ethical issues. I think it's in large part, that it will have mostly beneficial effects of delivering better health for people to have higher standards of living for more of their lives.

Swanson: But they're going to be some challenges and some temptations that maybe we don't want to go with. But I just say that, in the next 15 years, I think we're going to see huge advances in health care, which is one of our most challenged industries as far as productivity and one of our largest. We're supposed today at 17% of the economy, it's supposed to grow to 20%. And it's just far too expensive for what we're getting out of it. But I think there are a number of ways whether it's through personal technology, we're finally about to see the real fruits of the genomic and proteomic revolutions, it's going to, reduce the cost of drug development.

Swanson: So, I'm really excited about healthcare, again, because it's such a large part of the economy and so expensive, that also means it's a huge opportunity.

Beckworth: Well, our time is up, but this has been a very refreshing and positive message. A productivity boom is on the horizon and we have great social change to look forward to, but great opportunity as well. Our guests today have been Michael Mandel and Bret Swanson. Michael and Bret, thank you for being on the show.

Mandel: Thanks very much for having us.

Swanson: Thank you David.