Reading · Mortimus · RSS Amplifier

Editors note: Welcome to issue #3 of Transcriptions. Today, we’d like to present a conversation between hosts Tano Santos and Michael Mauboussin and their guest James Bessen. As always we couldn’t find a solid transcript, so decided to take up the mantle ourselves.

James Bessen shares his insights on the evolving landscape of competition and innovation, the impact of proprietary software on industry disruption, exploring the significant shift from mass production to mass customization facilitated by technology and much more.

Key Topics:

Welcome to Season 10 of Value Investing with Legends (0:20)
Exploring the concept of barriers to entry in value investing (1:05)
James Bessen discusses technology's economic impact (3:15)
Analysis of the decline in economic disruption despite technological growth (5:10)
Impact of proprietary software on industry stability and market concentration (6:35)
Historical perspective on technology's influence on the economy (8:25)
The role of proprietary software in shaping competitive dynamics (10:30)
The transition from mass production to mass customization in retail (12:05)
Walmart’s evolution and strategy in retail sector dynamics (14:20)
Discussion on the concept of skill premium and technological advancements (24:55)
Artificial intelligence’s influence on market dynamics and job skills (26:10)
Examining the pace and implications of innovation diffusion (31:00)
Regulatory challenges and tech dominance in the market (34:10)
Concentration of innovation and its economic implications (36:20)
James’ book recommendations (43:10)
And much more!

Mentioned in this Episode:

Tano Santos (00:00:01) - Welcome to a new season of the Valley Investing with Legends podcast. My name is Tano Santos, the Robert Heilbrunn Professor of asset management and finance at Columbia Business School and the faculty director at the The Heilbrunn Center for Graham & Dodd Investing. I'm here with my co-host Michael Mauboussin, an adjunct professor at Columbia Business School and a faculty member at the Heilbrunn Center as well. Hello, Michael. How are you doing?

Michael Mauboussin (00:00:24) - I'm great title. How are you? Hope you had a good break, by the way.

Tano Santos (00:00:27) - Yeah, yeah, it was a good spring break. Back in the classroom with the students and doing the value investing with legends, talking with some great names on value investing. So having a bit of fun there.

Michael Mauboussin (00:00:36) - That's fabulous.

Tano Santos (00:00:37) - So Michael, we're trying something in this season of the Value Investing with Legends podcast that is new. And that deviates a bit from what we've done in previous seasons. We want to widen the scope of our discussions, keeping in mind our listeners who are portfolio managers, investors, analysts, students, alums, why don't we spend some time explaining what we're trying to do before we introduce today's guest?

Michael Mauboussin (00:00:59) - Sure.

Michael Mauboussin (00:01:00) - A central idea behind what we call modern value at the Harvard Center is the idea of barriers to entry as a key ingredient in understanding value. For example, you have to worry about earnings growth if and only if the business operations of the firm are protected by barriers to entry that allow the firm to earn a return on invested capital above the cost of capital. Barriers to entry are crucial to investors today because when combined with a large total addressable market, a firm can reinvest capital at high returns and generate significant shareholder value and investor returns. This was Charlie Munger central insight, and he was one of the first people to understand this point and all of its profound implications.

Tano Santos (00:01:39) - Yeah, that's right. And we want to understand this issue of barriers to entry and total addressable markets in the context of technology broadly understood. The conversation today is dominated by the remarkable appearance of artificial intelligence in our lives, including its enormous promises and dangers. But we're also interested in what it's going to do to markets, regulation and competitive dynamics.

Tano Santos (00:02:00) - We plan to have four conversations on the topic. They will most certainly have some overlap, but each will look at the issue from a different point of view or emphasize different things. Today we will look at the issue from a social, economic, and regulatory point of view. To highlight just one aspect of this. We are recording this podcast barely a week after the announcement of the Department of Justice landmark antitrust lawsuit against Apple. We'll then have a conversation that tackles the issue from the perspective of global regulation, with an eye on global competition and the many issues there. Finally, we will have a conversation on competitive dynamics and technology using the framework that we all teach here at Columbia Business School. So that is going to be a very pedagogical podcast. We'll finish the series with a conversation with a tech investor, who will hopefully help us distill all the insights in a way that can inform how an investor should navigate this period of technological innovations.

Michael Mauboussin (00:02:51) - Absolutely. Technology is obviously been central to our lives and economic dynamics for at least 300 years since the invention of the spinning Jenny and the steam engine.

Michael Mauboussin (00:03:01) - So hopefully we'll be able to bring some of those lessons of history to bear on this conversation. So we ready to introduce our guest today, Tano.

Tano Santos (00:03:07) - Absolutely. Let's go for it. So our guest today is James Bessen, executive director of the Technology and Policy Research Initiative at Boston University. And in my mind, one of the most thoughtful observers of our current technological moment. He has published widely and in particular two books that I like very much, Learning by Doing and his most recent one, The New Goliaths, which are wonderful reflections on the state of technology, its impact on the economy, regulation, and the law. I recommend them to anyone who's interested in the current technological landscape. Jim Bessen, welcome to the Value Investing with Legends podcast.

James Bessen (00:03:42) - Thank you for having me. I'm glad to be here.

Tano Santos (00:03:44) - We're delighted to have you here. So I want to start in a perhaps surprising place, which is your observation early in the book, in the nucleus in the last 25 years, if at all, the pace of disruption in our economy has in fact gone down, not up.

Tano Santos (00:03:59) - I'm sure many people would find this surprising, given that technological change seems to be everywhere around us. But in the very first chapter of your book, it is entitled in fact, ‘Disruption Lost’ and documents that in fact there are disturbing trends across industry. So why don't we start that? You tell us a little bit about those trends and we'll take it from there.

James Bessen (00:04:16) - So first, I should qualify exactly what I say, because in a very broad general sense, disruption is not diminished. Certainly things change. We do things in very different ways. We communicate. I mean, the cell phone, the smartphone is changes behavior in all sorts of ways. But in the very specific sense, this sense that Joseph Schumpeter talked about long ago, the leaders in industries have been disrupted less and less. So this is sort of a key metric as to how much what Schumpeter called creative destruction is going on is to the extent that you have incumbents, leading industries, and are they getting disrupted by new firms coming along with new and different ways of doing things, often based on new technology? This is something we can measure.

James Bessen (00:05:06) - So we looked at what is the likelihood that a firm that is in the top four of its industry, in terms of sales, will be in the top four of its industry a year later or four years later or whatever. We measure it a variety of ways, and the pattern is striking regardless of which way we measure it, which is the rate of disruption was rising in the 70s and 80s and 90s, but in the late 90s, early 2000, it peaked, and it has gone down sharply since then. The rate of disruption is about half of what it was back then measured in this very specific way. So that's very telling. I mean, you know, you've got Clayton Christensen talking about how disruption is taking over and a whole little mini industry of business consultants emphasizing disruption and how we have to look out for it. And literally there's a lot of leading firms in a lot of industries that have figured out some way to make disruption less of a threat to them. We looked at this.

James Bessen (00:06:09) - We looked at what was at least associated with it or what was causing it. We see some parallel trends. So industry concentration has gone up, meaning the share of sales going to the top four firms has risen broadly across many industries, not tremendously, but significantly, which says again that the biggest firms are getting bigger. We are able to relate these changes to technology that it turns out that it's in industries that invest more in proprietary software. Let me back up and talk about software second. We tend to think of software being the prepackaged software we get from Microsoft. That's great, but it turns out that the majority of software investment that firms make is actually their custom software. They develop it themselves, and they spend billions on it. Walmart spends 10 billion a year on developing its own internal software. So it turns out that industries where a lot of proprietary software is being developed tend to be the ones that have seen their industry concentration go up, and they tend to be the ones where the disruption has lessened.

James Bessen (00:07:20) - And so that's set off the set of investigations that I wrote about in the book, but also some of our ongoing research.

Tano Santos (00:07:28) - Can I ask a follow up on Michael? There's so many things to unpack in Jim's answer. So you mentioned Schumpeter, and I want to take a trip down ‘History of Economic Thought Lane’, a little bit. And one of the things that Schumpeter was worried about in his book, Capitalism, Socialism and Democracy, was precisely this idea that eventually the companies were going to develop these massive R&D platforms, that innovation would become routinized inside corporations, that in a way, it's a very wasteful process, that of innovation, and that rationality will call for these R&D to take place inside these large firms. Are we seeing the manifestation of that schumpeterian fear?

James Bessen (00:08:14) - Depends exactly how you interpret it. But what I have to say is, yes, R&D is surprisingly concentrated. 90% of it is done by 200 some firms. We tend to think that this is something it's across the economy. Really. Very few firms invest in R&D.

James Bessen (00:08:32) - They tend to be large firms, not all large firms though. However, part of it of course depends on industry. But even if you take industry into account, there's this huge lopsided effect where it's dominated by a few firms. And here's the interesting thing. The price of performing R&D has gone down, and partly because the price of software has gone down. R&D consists of 3,040% software expense. Even as the price has gone down, the share of firms that are doing R&D has not increased. In fact, it's decreased a bit. So this is surprising. It is saying very much that the innovative effort is highly concentrated, and dropping prices aren't helping the problem. They may in fact be worsen it. The same is true if we broaden the concept of R&D to include custom software development or software more generally.

Michael Mauboussin (00:09:23) - I'll just follow up and then I do have a follow up question, but I don't know exactly what the R&D number is for corporate America. It's probably was a 600 billion or 650.

James Bessen (00:09:31) - Billion close to a trillion now.

Michael Mauboussin (00:09:33) - And then Jim, in your book, you write that proprietary software spending—I just want listeners to understand the context—it's about 235 billion. That was a few years ago. So probably a little bit higher than today. So it's a very, very substantial sum.

James Bessen (00:09:46) - Yeah, I sort of alluded to it that so much of the software investment that's going on is about two thirds of it is firms investing in software for their own use, basically not for selling it on the market. We tend to think of the prepackaged software. There's this other thing going on, and it's a very different animal. It's not something that goes on the market. Most of it doesn't have a market price of any sort, or Walmart doesn't sell or license their internal software. They use it to gain competitive advantage, and it's a tremendous factor in gaining competitive advantage these days.

Michael Mauboussin (00:10:22) - Let's jump into that a bit. You emphasize the issue of competition through complexity. So how large firms, including companies like Walmart, essentially leveraged the complexity only available to them through their software to obtain competitive advantage.

Michael Mauboussin (00:10:35) - So can you walk us through that idea in particular, and more specifically, the role of intangibles? I mean, I guess role of intelligence in general. You mentioned software and maybe some concrete examples of that proprietary software. We've alluded to Walmart a couple times, but how that proprietary software allows these companies to decrease the rate of disruption or sort of keep competitors at bay.

James Bessen (00:10:55) - Let's start way back. Think the industrial revolution and technology then was very much about quantity produced rather than variety. Now, that's not to say there weren't important quality improvements. There were. It's not to say that variety wasn't important, it was. But if you look at something like you talked about the spinning Jenny, talk about the textile industry, [which] in the US grew by making highly standardized cloth and great volumes. In Britain, it was a little different. It was more specialized. They did some fancy goods, but it was very clear it was in the US that the real productivity benefit and the real growth occurred at a much larger scale.

James Bessen (00:11:36) - So it was very much this mass production model, and that got built on an emphasis as later in the century, new technologies, the steam engine came along, mechanization came along. But the key to this was that these were highly standardized products. So they were able to produce a lot of it at low cost. And that was a tremendous advantage. But of course, always the problem with that sort of model is it doesn't meet individual or idiosyncratic needs. I may want something that isn't being offered. And the famous story was all the early Model T and all the Ford cars were black. And Henry Ford said, you can have any color you want as long as it's black. General Motors came along and started offering cars in color, and it turned out that he was his very modest change, and it turned out to be a significant market winner. But the basic tradeoff was there between mass production and sort of lack of quality and variety. Software is unusual in that it is very modular and it lends itself to producing varieties very cheaply.

James Bessen (00:12:44) - So you talk about Walmart in part, what its software does is allows the chain store to handle a very large number of items to put on shelves and handle them in a way that not only tracking them on the shelves, but tracking them through the cash registers, the whole logistics flow of getting them from suppliers to warehouses, the trucks to stores. All of this being done through software, can provide a tremendous advantage and an ability to handle a large variety of products. So 100 years ago, it was A&P and the big chain stores standardized what they were offering so that every A&P had the same products and they could gain efficiencies by having warehouses stuck in the same products and, and all that. Here you've got Walmart. Now using software is able to handle a huge amount of variety. This is the complexity I'm talking about. But you see it not only in retail. You see it in manufacturing. So cars now are software based. All sorts of features of cars are handled in software. And so it provides great flexibility in terms of tweaking the features, adding features.

James Bessen (00:13:57) - Software itself is the whole issue of software bloat, that there are feature wars that if you look in the 80s when software personal computer software was booming, it was all about each year, every company would come out with a new offering that would have a whole set of new features, and so they could add features quickly and readily. In finance, the banks can offer all sorts of different credit offerings and tailor them very quickly and market them, target market them. So all of this is a way that we're getting away from the mass production model into a mass customization version, where things can be tweaked and tailored to the needs of individual customers, and that's a huge gain. It's very hard to estimate how huge, but it's something that gives these firms a big advantage. The credit card companies that use these big information systems dominate the credit card industry. Walmart grew from 2 or 3% of the market to 50 some percent of the market in general. Merchandise tremendous benefits.

Michael Mauboussin (00:15:02) - So, Jim, let me ask you, as a follow up, Michael Porter, obviously famous strategy professor, used to talk about generic strategies, low cost producers, what you just described.

Michael Mauboussin (00:15:12) - So kind of going to the mass market, like you said, uniform but cheap and then differentiation. So sort of customization. So he had this very strong admonishment don't get stuck in the middle. Don't try to be a low cost producer and try to differentiate. But it sounds like what you're saying the last say quarter century, this is precisely what companies are able to do in a way they couldn't before. And it also sounds like this is a bit of a positive feedback. In other words, I gain a little bit of advantage through technology. It allows me to have more customers, which gives me more data, which allows me to spend more and, and sort of this positive feedback. So is that a fair characterization? Do we need to set aside the old Porter admonishment? Is this something that's going to be here to stay? How do we think about that?

James Bessen (00:15:49) - Yeah, I haven't thought about it really from the business strategy point of view. But I guess, I mean, it sounds right.

James Bessen (00:15:55) - The difficulty is think about Walmart's small competitor. Even not so small. So Dollar General is a big chain. Where Walmart will have 140,000 stock keeping units in its supercenter stores, Dollar General will have 10 or 12,000 much, much, much smaller selection. But what Dollar General does is it sprinkles its stores geographically. So it's particularly in rural towns that sort of area, it's getting closer to some consumers. So it has a geographic advantage. And Dollar General spends nothing on proprietary software. They use some off the shelf software, but it's a distinct product in the end. Well, it's going to be interesting to see how it all plays out. So there are some strategies where you can stay out of the middle, and maybe there's some virtue to that. And it's also clear it would be difficult for a Dollar General to increase its product variety without making huge investments in a change to its business model.

Tano Santos (00:16:55) - So I guess the question I have about this, and we're completely off script, which we knew we were going to be off script because we know the conversation was going to be wonderful and take a similarly different directions.

Tano Santos (00:17:05) - But if I think about a company such as Walmart, if I think about the evolution of profitability, operating margins in that company of ROIC, for instance, going down steadily, my intuition for that has always been that, yes, they spent a lot of software. They have a unique logistical ability. This ability is not one that improves its competitive standing, because it can be replicated by the other large retailers. And Walmart has been notoriously bad at competing with large foreign retailers, and they are getting out of the international market. The point is that you think that only large companies are going to be able to sustain the type of R&D expenses that keeps them as market leaders, even though it may not translate into profitability because they're always subject to the threat of entry. Is that the way you see it, Jim?

James Bessen (00:17:59) - I'd be cautious about extrapolating from the Walmart, the example, Walmart's margins.

Tano Santos (00:18:05) - Right.

James Bessen (00:18:06) - Let's see. The difficulty is and as I related earlier, we are seeing a concentration of R&D and a concentration of software investment.

James Bessen (00:18:15) - So it's not like it's diffusing elsewhere. Absolutely. The ideas that Walmart develops diffused over time, in fact a lot of them diffuse to Amazon. Amazon poached Walmart's top executives. They had a lawsuit over it. But eventually Amazon was able to gain a lot of critical knowledge that had been developed by Walmart and Walmart's people and use it in its own development. So these things happen, diffusion happens. But I think the important difference is it just this happening at a much slower pace than it used to happen.

Michael Mauboussin (00:18:51) - Well, Jim, I think we should dwell on diffusion for a moment. And I think you've got a delightful story in the book about General Motors developing automatic transmission and how it took about ten years for that to diffuse to all the other manufacturers. And that was with a world war in the middle of that time period. And you also argue that diffusion seems to be happening a lot slower. So Walmart had happen, but it took some time. So what is the impediment to diffusion these days? Why does it not happen faster?And is there anything anybody should be doing about that, whether it's regulatory or anything else?

James Bessen (00:19:21) - The answer to the last question is, yeah, but it's complicated. I think there are two differences. One is that there's less incentive for firms to promote diffusion themselves. So General Motors actually licensed its automatic transmission technology to some of its rivals, including Ford. Not all lines and not all models, but and it makes sense to license your technology if you of course, earning licensing fees. But if it's the kind of thing where automatic transmission is used, it grows the market. More people will buy cars if they don't need to learn the standard transmission that increases the size of the market and so it's not a zero sum game. You licensing it to Ford. If we end up with a zero sum game where every customer that Ford is able to get because would have otherwise gone to General Motors, then that doesn't make any sense to license. And with the new technology, because it's so much geared towards differentiating products rather than expanding the market overall, it often doesn't make sense to license these products.

James Bessen (00:20:28) - We've seen some interesting examples where Amazon is. In fact, I think an interesting example that I elaborate in the book where they have licensed or made available or opened up in some way parts of their technology, and I think that's a very beneficial thing. But I think the general tenor is one where firms are reluctant to do that. They're understandable reasons for that.

Michael Mauboussin (00:20:50) - How do we understand Amazon Web Services in that context? So in a sense, that was a proprietary technology that they opened up. So they're charging people to use it. And it seems like it's almost like a middle ground here versus the original General Motors Ford thing versus not doing anything.

James Bessen (00:21:05) - I think it's the same sort of logic and similar to the logic of IBM opening the mainframe, the mainframe software back in 1969. Amazon absolutely had a competitive advantage in its internal IT infrastructure, and it knew it had a competitive advantage. It was getting requests from places like Target, where they wanted to have access to those tools, and they realized there was a big market out there for customers.

James Bessen (00:21:31) - And the critical question is, by opening it up, are they going to expand the size of the market? And they hugely expanded the size of the market. Cloud computing is just far beyond what anybody imagined it would be. I think one of the difficulties with these sort of instances, it's very often hard to predict, and even the insiders don't have a very good clue to it with IBM. IBM commissioned studies to see whether they should open up their architecture and allow third party software vendors. They had two internal studies, and they were both kind of mixed and they had no clue. And it turned out to be one of the most profitable things they ever did. It's about expanding the size of the market. And I think in so many industries, we're in that place. Company leaders don't see it. And everything they've been conditioned to worry about and think about blinds them to this possibilities of those opportunities. It is a case where smart regulation can actually improve things. And in fact, the antitrust authorities did play a role in coaxing IBM to open up.

Tano Santos (00:22:34) - Can I add one thing that, you know, you can throw into your discussion, Jim? One second if you can. I mean, I think the reason why many times companies do this is because they want to create an ecosystem of software developers who are going to build complementary applications that will reinforce something that the company that is opening up the system will benefit from, and would allow them to capture the rents associated with a growth in the market. Because there's a particular technology or infrastructure, in the case of AWS, which they have a monopoly. I think, for instance, about the open software movement, so to speak, Linus obviously immediately comes to mind. It seems to me that this is evolving, and we're going from an open to a closed system where the large companies, if I can connect it to your earlier comments, are beginning to say, well, I can do it myself, you know, I will create these large markets myself, and I will give some incentives for developers to develop software.

Tano Santos (00:23:30) - But I extracted enormous amount of rents and I keep it very close. I mean, in a way, the Apple lawsuit has some aspects of that to some extent. Is that our technological moment? What do you think, Jim?

James Bessen (00:23:40) - It's very clear. There are degrees of open. Open is in the details. And this is why it's such a difficult regulatory problem. You can look at Apple. There's no doubt that the App Store was a huge opening. And it was another one of these cases where it expanded the market dramatically. It was beneficial to lots of people, but it's also clear that Apple put its thumb on the scales in a few places and twisted it to work in a way that maybe isn't the best for society, but is good for Apple's profits. Maybe there is a regulatory role there, but it's a difficult one. I think it's going to be very interesting how it plays out, but this is, I think, one of the big challenges that we are facing over the next couple of decades.

James Bessen (00:24:22) - We're going to see more of that. I think we're going to see more companies opening up in some way. At the same time, there's going to be this tendency to open up with restrictions, set up walled gardens.

Michael Mauboussin (00:24:34) - Jim, I shift gears a little bit, just talk about the some of the social, political impacts of these innovations and one of the key findings in economics, and you've written about this is the increasing skill premium. So workers with high degrees of education, since the sort of mid 1980s seem to earn consistently more than less educated workers. And that's been interpreted as a skills based technological change that's being reflected. What happens with that going forward? And I think one of the interesting notes is that the ‘New Goliaths’ was published, I think early half of or middle part of 2022 and late 2022 when ChatGPT hit the scene, and AI—which had been around for a long time—all of a sudden became much more part of people's awareness, both businesses and investors. So how does AI fit into this whole thing? Is it something that just strengthens the strong guys or it opens the door? Now for shuffling a little bit, given this lack of competition?

James Bessen (00:25:29) - So there's a couple of questions there.

So first off, one interesting and important thing is that the skill premium stopped growing. And depending on exactly which numbers you're looking at, it is maybe ten years ago now, I think I wrote about that in ‘Learning by Doing’ so, it doesn't mean it isn't important, but it's not something that's necessarily going to go endlessly into the future. The AI story is pretty interesting. My first inclination is that in many ways, AI is going to be just like IT, but on steroids. We still have to see how that all plays out and it can play out in different ways. We're seeing a lot of demand, for if you look at demand for data scientists or data analysts, or what you're seeing is companies want to hire marketers who have data capabilities, you're looking at hybrid skills. So increasingly you're looking at different sorts of skills that are learned on the job that require some experience. That's a whole problem, an issue of itself I talked about in ‘Learning by Doing’, but also it's not clear how it's going to skew across educational groups.

James Bessen (00:26:38) - So in some studies they're finding for instance with the study of software developers, that AI works to improve the capabilities of the weaker software developers and doesn't affect the top guys so much. In some cases, it may even deteriorate the performance of the top guy, the most capable guys, so there may be more opportunities. David Autor argues there are more opportunities emerging for the mid skill occupations to benefit from the technology, but it's very much a matter of combining the occupational skills with experience using these data systems. There's a whole training and educational issue that we're confronting.

Tano Santos (00:27:24) - If I can ask another big question on this, and I just cannot resist the temptation of asking you this, which is where do you stand on this debate that we currently have on economics about the future of productivity growth? And where do you stand on the Robert Gordon's thesis in The Rise and Fall of American Growth that the big innovations: electrification, indoor plumbing, vaccines, things like that are behind us? Those are big productivity boosters and the smartphone is fine and it's fun, but it doesn't compare to those type of technological innovations.

Tano Santos (00:28:02) - So we shouldn't be expecting the type of productivity growth numbers that we were able to get in the past, which, by the way, just to be clear, the data seems to be telling a story that productivity growth or total factor productivity growth has been slowing down for quite a while. There are obviously increases here and there, but overall the trend doesn't seem to be positive. Where do you stand on this debate?

James Bessen (00:28:22) - So first off, you have to think a little bit about what productivity measures. It was very much something that grew out of the industrial economy. You think about an economy of mass production. You want to measure the output per worker. That's the labor productivity measure, and you make adjustments for capital investment and such. But it's a very clear idea. If you're measuring a commodity where there's a single output and you can measure quantity clearly.

Now, it's been known for decades that when it comes to the sausage making of the productivity statistics, quality is a big and complicated issue. But if you now think that all of a sudden we're into a world of huge product variety exploding, the challenge to actually measuring productivity is immense.

James Bessen (00:29:09) - We're not able, I think, to measure the benefits of this new direction of technology in the same way. So if we think of productivity being just the measure of commodity outputs, then I think Gordon is right in a sense. But I think that's a limited view that in a sense we've shifted to a different kind of technology, a different kind of output that we're not measuring and not capturing very well. It's not so clear that we're in any sort of decline, but I think it's very hard to know. We're sort of at the limits of what can be measured.

Tano Santos (00:29:42) - Can I add a note to this? So there's an interesting concept right now. There's a beautiful paper by two scholars at the University of Chicago, Esteban Rossi-Hansberg and Chang-Tai Hsieh, on The Industrial Revolution in Services. And they make a very interesting point. They said, well, these technologies are allowing for economies of scale to work their wonders in services, which is the type of economies of scale that we're used to in manufacturing are now operational in services.

Tano Santos (00:30:08) - And they mentioned things like healthcare diagnoses that we can do imaging and diagnoses on centralized databases that stay far away from where the actual care is being supplied. And all of a sudden, the idea that we had that the provision of services has to be local. Well, to some extent, yes, because you still need a nurse to perhaps, or a doctor to do a procedure. But to some extent, many of the steps that go into the provision of health care can be subject to those economies of scale that until now, we thought were the domain of manufacturing. So that can be a big source of productivity improvement.

James Bessen (00:30:49) - I mean, this is in a sense the same thing I've been talking about. So the the big example they have is in fact retail.

Tano Santos (00:30:53) - Yeah, absolutely.

James Bessen (00:30:55) - There is a benefit to be able to go to a store that has 40,000 different items on its shelves, and I know I'm going to have a much greater chance of finding the things I need, and has even greater benefit if I can go online where there are millions of items that I can buy.

And so the nature of retailing has changed in its ability to meet consumer needs has changed dramatically. But we don't know how to measure that.

Michael Mauboussin (00:31:19) - We want to turn to M&A and regulation and antitrust. Before I do that, let me just summarize this whole discussion. What you're saying is strong companies last, say, 20 years have become more entrenched, in part because of this proprietary software that's not diffusing. It's harder for smaller companies to break into the ranks of the big companies, the big companies we know. And people talk about the Magnificent Seven, and these are very profitable companies. So they're growing at actually remarkable rates for their size, and they're quite profitable. Is this all a problem that needs to be fixed by somebody, or is it just a natural outcome from the way the world goes?

James Bessen (00:31:51) - Well, I think the problem arises. I want to be careful not to focus on big tech and gaffe and whatever. The problem that I'm talking about is a much broader segment of the economy that includes banks and includes retailers.

It's not just the big tech companies. That may be the most dramatic examples. I'll leave it at that. The problem comes in the long run that if you've got Walmart, which was at one point the up and comer, it displaced Sears, it displaced Kmart. But are its margins going to deteriorate over time. And is there a problem with new firms entering with new ideas and their ability to grow? So we're seeing lots of evidence that there is such a problem like that, that the future may be in trouble because nobody's got a monopoly on the next innovative idea. Right? You need to have a diversity of sources for innovation to be very healthy. And when you increasingly concentrate innovation in a few companies, ultimately you're throttling the potential gene pool for which new ideas can come from. So there is, I think, a very significant downside possibility. I think there's some other social implications for that sort of economy. One is you're talking about very high concentration, and I think there's a greater risk of political influence from dominant companies, especially in a political system where money matters.

James Bessen (00:33:22) - The other is we've tended to look at small companies’ entrepreneurship as being a major source of social mobility. And if we throttle the ability of productive small companies to grow, they may have a great innovation, they may have a great idea, but if they can't grow well because they're competing against firms that have a broad, broad range of products to offer, then we're limiting an important source of economic mobility. The evidence is quite clear that innovative startups are not growing as quickly as they did in the past, that I think is maybe the most critical downside.

Michael Mauboussin (00:34:02) - So notwithstanding that, I mean, does it seem like the large technology companies are in the crosshairs of, for example, the antitrust folks? We're recording this shortly after the Department of Justice filed its lawsuit against Apple. So this issue of antitrust enforcement is going to be going forward for the companies and for investors. What might that might mean for technology diffusion, growth and profitability of these big technology companies, in your view?

James Bessen (00:34:27) - I think to some extent, the antitrust authorities are focusing too much on big Tech and not on the broader problem, which is quite significant and affects much more of the economy.

Big tech is 2% of the economy, maybe 3%. And in many areas it's quite clearly highly competitive. That doesn't say there isn't room for antitrust regulation. I think there really is. We talked about Apple, and I think there's some pretty clear cases where Apple put its thumb on the scale and tilted the way that it runs the App Store in its favor, but the App Store overall is a great idea and a great benefit. Similarly, Amazon has been attacked, and you look at the way Amazon has opened up its marketplace and opened up AWS. And those are the kinds of things you want to encourage. Now, there certainly is room to criticize how it's done things within those platforms, but we don't want to discourage that idea because you've got most of the economy. Think about the big banks, and they've stymied fintech companies from getting customer data for decades now, even though the laws are there to supposedly encourage them to do so or even force them to do so. We've seen very lax regulation in that way, and I think it's pretty clear comparing fintech growth in the US to other countries where they've now started to get the banks to to share data.

Tano Santos (00:35:58) - So you were talking about fintech and the banks. If you think about what mobile applications have done to competition in the banking industry, and we have some very good research done on this. Now, what you see is a lot of entry of large banks in small markets as a result of the local banking market, which typically was very concentrated because it was dominated by the local bank, who knew everyone in that particular county say is now more competitive, and some of the welfare or gains associated with that technological innovation are flowing to customers who now have more options for deposits, for perhaps mortgages. So you see beneficial aspect of technology in terms of increasing competition.

I think the concern that if I can go back one second, I know that the Apple lawsuit is very big in our minds these days, is the type of actions that are taken to prevent competitors to, say, pair their watches with the iPhone and say that's their concern that the Department of Justice has.

Tano Santos (00:37:04) - So they are trying to be mindful that, yes, of course, I want to encourage this great company to supply goods that people seem to love. They clearly love their Apple products while at the same time preserving options and choices. Do you think that in that sense, that after a period where perhaps antitrust was less active, we're entering a period where it's going to be more active and more targeted? If I look at the Apple complaint by the Department of Justice, which I recommend everyone to read, at least the first few pages seems very targeted. Something about texting, for instance. You really want to preserve the quality of texting and interoperability, which seems like something important. The fact that you can pair your watch and that your watch works equally well as an Apple Watch when it's paired to an iPhone, and so on and so forth. This idea of interoperability to allow the ecosystem to grow in a competitive way.

James Bessen (00:38:02) - I agree interoperability is very important. I'm less hopeful that regulators can at this level in the weeds, you know, get them the answer right.

Tano Santos (00:38:13) - In a way, yes. Get it right.

James Bessen (00:38:14) - Yeah. Because you're talking about very detailed things. It's easy to read some of the critiques out there about some of these things. Apple itself is making changes partly in response to the regulatory pressure. These things seem to be relatively marginal to me. They're not holding back big opportunities. I think with fintech and the banks, the banks are holding back some real beneficial activities that would benefit consumers. It's not clear to me the color of the balloons in messaging apps between iPhones and Android phones is a big issue that matters very much to social welfare. That doesn't say they shouldn't regulate it, but this is pretty small beans. The courts tend to give firms a lot of leeway to control the terms, especially where they can make an argument, which Apple clearly makes, that this is for privacy and security and ease of use of the consumer. They've done an excellent job at providing those things. And yeah, if they cross the line a little bit in one place or the other, maybe the antitrust authorities should enter into discussions with them about that.

But it's not clear they want to break up Apple over something like the color of the messaging balloons.

Tano Santos (00:39:30) - No, no, of course, that's exactly what I meant. What I meant by that is that it seems to me directed or a theme that I detect on the complaint is this interoperability issue. They're not threatening to break up Apple or anything like this, but really to facilitate the ecosystem to grow outside the domain of the R&D platform of Apple to other people who may want to supply. An interesting thing is, for instance, this thing of gaming on the server, you access a game application on your iPhone, but you can play it on the server. And in that sense, Apple acts as a corridor. If you want to think of an image that that seems to me like a reasonable thing, it's going to be harmful to Apple shareholders in terms of profitability. Seems to me a reasonable thing to allow customers to enjoy. Having said that, by the way, I wonder whether today the Meta acquisition of WhatsApp or Instagram would go through.

Tano Santos (00:40:29) - It seems like a different environment all together. Can I ask you then, a broader question, if you were to take a step back with your concerns on diffusion, the importance, by the way of experimentation to get the innovations in the right? I think something that you discussed at length in Learning by Doing. Do you see or what would you would like to see in terms of intervention in these markets? What kind of regulatory repairs? Would you like to see to make sure we remain innovative?

James Bessen (00:41:02) - Yeah. Let me first say express humility, which is I don't have very good answers. And I think in reality, we need to stumble through in the direction of improving openness and interoperability. But it's going to take decades for us to figure out how to do it well. There are clearly some opportunities where antitrust authorities, I think, can take some steps to open things up. What I would like to see is for them to focus on the big issues. It worked to encouraging firms to open up key aspects of their platforms or their data, or in some cases, force them to do so.

For instance, if to resolve an antitrust lawsuit or to permit a merger to go through. I think all of those things are sort of in the domain of what the law allows for.

Tano Santos (00:41:52) - So we're getting to the last part of our segment. And, Jim, we always ask our guests two questions to end the conversation. Mine is what keeps Jim Bessen awake at night with excitement or worry? What do you think is that thing that we should keep an eye on?

James Bessen (00:42:12) - Well, if I've drunk too much coffee during the day. But that's another story. That's another story…

We're going into this new era and we're stumbling into it. We have this big challenge that we have this tremendous technology that's able to meet all sorts of individualized needs on a scope and scale not seen before, and yet we just haven't figured out how to really do it well, so that we're keeping things innovative and the benefits are spread around. And we're not increasing inequality, we're decreasing inequality and opportunity. That seems to me to be the challenge.

And it's both a fear that things are going to go horribly wrong, and we're going to see a choking off of technological progress or an excitement that there's just a lot of opportunity to improve things and make the transition. So we have a society that's able to use the technology in a good way.

Michael Mauboussin (00:43:09) - So, Jim, what are you listening to these days, or what books are on your nightstand that you would recommend to our audience?

James Bessen (00:43:16) - For my own personal reading, I tend to either read totally escapist stuff or highly technical stuff, but right now, actually, I'm revisiting The Complete Sherlock Holmes, which is quite a lot of fun. A book I can recommend very easily is Thomas Philippon's The Great Reversal. It's a compliment to some of what we've been discussing. I also tend to think back economic history has always been very important to my thinking about things, and I would encourage anyone to read about the Industrial Revolution. And a great book is Joel Mokyr’s The Lever of Riches. It's an old book now, but it just tells this story through history of how technology changed society and how innovations were encouraged or discouraged over time and different ways.

And Joel's got a bunch of interesting books, but this is probably a great introduction. A third one I read a year or so ago, an economist who wasn't very well known today outside of economic development circles is Albert Hirschman. And there was a great book, Worldly Philosopher: The Odyssey of Albert O. Hirschman. Well, one, he had an incredible life, the resistance to the Germans in World War Two and prior, but was always a heterodox thinker. And he's an interesting person to read because he comes with a sort of new set of concepts in terms of thinking about things. The biography is a great sort of overview of a lot of his thinking, and it's just kind of refreshing to they use the phrase outside the box these days. But I think, you know, he's very much a distinct thinker who I find it stimulating to hear his new ideas and think about them.

Tano Santos (00:45:01) - I completely agree that reading about the Industrial Revolution in particular seems pertinent these days.

I think it's going to be quite important to take the lessons from history and a few people out there who do it as well as you do. So Jim, the author of Learning by Doing, The New Goliaths—two books that we very much recommend to all our listeners. Thank you so much for coming to the investing podcast. Thank you.

If you enjoyed this transcription, you’ll like our previous one, a conversation between Qasar Younis and Bilal Zuberi:

Transcripts

Building Great Companies — Fireside Chat with Qasar Younis

June 5, 2024

Editors note: Welcome to issue #2 of Transcriptions. Today, we’d like to present a conversation between Qasar Younis of Applied Intuition, and Bilal Zuberi of Lux Capital. We discovered Qasar recently, and find his focus and clarity of thinking to be admirable. What follows is a lightly edited

No posts

The Tech Paradigm — a conversation with James Bessen

Building Great Companies — Fireside Chat with Qasar Younis

Comments