Luohan Academy

Why Do We Need New Measurements for Human Progress? | Frontier Dialogue

Nobel Laureate, Stanford University Professor, and Luohan Academy Academic Committee member, Michael Spence, demonstrates why we need to measure the new economy, why it is important, and its future implications.

This Dialogue intends to provide a forum for prominent scholars, policymakers and industry experts to exchange views and collaborate on possible new and better measurement approaches.

Luohan Academy has been in close collaboration with some of the world's foremost social scientists, policy experts, and business practitioners to document and analyze society's and the economy's ongoing digital transformation worldwide. Our very first Frontier Dialogue focuses on the measurement of the new economy, a crucial step in the direction of better understanding the impact of digital transformation.

Transcript:

Michael Spence: And I guess I'd like to, you know, Erik and Dirk are gonna take us through some very interesting material that's important. But I'd like to stand back as I was trying to do a year ago when we were... or a year plus ago, when we were in and sort of, you know, think about, you know, the measurement issue in the context of what I think of as well-being. Whoops, now I can't move this. Maybe if I do that. Okay. So most of us, I think, you know, believe, I don't want to overstate this, that something quite fundamental is happening in our economies and that much of that is being driven by digital technologies and their applications in a wide, wide range of sectors.

So I'm not gonna read this list and it's not meant to be complete, but the point I wanna make very briefly on this slide is that the footprint in terms of impact of digital technologies is just enormous. It runs all the way from labor markets in the coordination of economic activity to how capital markets work to e-commerce and fintech, you know, to the issues that Erik is gonna talk about, free and low cost services, is gonna have an enormous impact on healthcare and education, and so it goes. And that means that if you really want to sort of have a holistic view of the impact of digital technology, you can't just focus on one of them, or you can focus on one of them at a time, but you won't get a complete picture.

And what worried me a year ago is we were focusing on traditional measures of economic performance and finding, you know, relatively little impact thus far. I'm sure, you know, Bob Gordon is going to help us understand that better. And on the other hand, we had discovered lots of problems that weren't anticipated several years ago, huge challenges and draining labor, you know, the labor force or rescaling it and data privacy issues and all kinds of things. And I thought, and I still think, to some extent we're missing the full range of the potential benefits, but because its footprint is so large, it creates a kind of what I call "radical multidimensional complexity" to this thing which is frustrating. It's very hard to bring it into focus. And there are now literally hundreds of indicators of various aspects of digital transformation and penetration and Dirk is gonna help us with these.

And these are really important because if we don't have yet the toolkit to get a complete picture of what's going on and the way social science and economics in particular works as we, we know we have a big puzzle and we have bits and pieces. And these data on penetration of internet-connected mobile phones and all kinds of things, you know, cloud computing systems, they're really important inputs. But I wanna just make it clear at least what my motivation is, and it isn't the long run objective, at least from my point of view is not the sort of understanding of every detail of what's happening in the digital economy and society. It's to know, ultimately the goal is to know, what the impacts are on various dimensions of human welfare.

And I underlined "various dimensions". I am not a fan of single measures. The only time single measures work reasonably well is when they move more or less in sync, meaning they're highly correlated so you can pick one of them. It works pretty well. And I've seen examples of that when we were working on early-stage developing countries, then it was pretty clear that GDP growth and income growth was a necessary condition for achieving a rather broad range of economic and social objectives in terms of progress. And so it was perfectly legitimate, so I'm not, you know, throwing it out. So I think, in a sense, the real challenge is, on the way through documenting digital momentum in various parts of the economy and society and in various fields, but to solve this problem of measuring impact.

And for that, I'm not going to be able to deliver on this, at least not in the near future. I think we need testable...we need theory. We need models or frameworks that capture the way in which the digital technologies that we're interested in affect the economy. And I mean, I could speculate about what that might look like. It's gonna have to have a network like structure. But except for some small inroads, we really don't have those models now. Not withstanding, you know, at least a little bit of progress and understanding the informational structure of markets. I think we're, we're in the early stages of getting there. Let me give you an example. This is in the spirit of multi-dimensional. And I'll do it very quickly. Some of you have heard this before.

So a couple of young researchers at Stanford work with top dermatologists in the medical school to use this very powerful general-purpose technology called image recognition to see if they can detect skin cancers. They have lots of different examples of this type and it worked quite well. I mean, there were subtleties about false positives and negatives that we don't need to spend any time on. But the interesting thing was, because the images could be taken with a camera, you could implement this on a very broad front. And so then if you just kind of understand roughly what happened and ask what's the welfare implication. Well, for people who live in Stanford, California, or London, or Hangzhou, it's probably the algorithms didn't beat the dermatologists. They kind of came close to a tie, I guess, it's the best.

The impact's relatively low. Nobody's gonna, you know, stay away from the dermatologists. But then if you kind of flip the switch, there's a huge populations that are frequently described in the literature as remote or low access. You know, they aren't anywhere near this aspect of primary care. And here, and in many other cases, you could have a very substantial effect and the effect is inclusiveness. So the dimension of well-being is health, the category is primary care and diagnostics, the impact is inclusiveness or reduction in dangerously late diagnosis, diagnosis in poor countries and regions. And the impact on GDP and similar measures is unsurprisingly, probably zero. And partly because of the pandemic economy, telemedicine is developing very rapidly in this, kind of fit into that category. This will overlap a little bit with what Erik, I think, is going to talk about, having taken a peek at his slides.

So I think it's important to understand the cost structures of these digital economies. They are high fixed sort of development/innovation costs and tend to have low to zero marginal costs. That means they're massively scalable. They have relatively low labor content in, if you just count numbers, I'm not talking about valuing the human capital and relatively small tangible capital content and huge intangible asset content and actually Erik and I, and Andy McAfee wrote about this a few years ago. But I think we need to come back to it. So we have lots of low cost or free services. We have superstar firms with substantial market power, especially associated with the digital assets and especially data. And one of the trends that we've observed, I gotta show you this in a minute, is that value creation in the capital markets is increasingly looking like it's the result of the buildup in intangible assets, not in all firms, but in the ones that are exhibiting abnormally large increases in market value.

And I know those intangible assets are low. We don't measure this very well. That looks like the digital ones and especially data are particularly important. The labor content is relatively low. So the value creation per employee is very, very high. And what we're starting to observe, and this is a trend that really worries me and in this scenario I want to do more work, is that employment, and to some extent, value creation using employment and traditional tangible assets are diverging from incremental value creation in the world, and much faster, and this has happened, much faster and kind of brought the issue into focus, although it had been observed before in the pandemic economy. And if you asked yourself who owns these assets, the answer is that the ownership is highly concentrated for fairly obvious reasons.

But one of the things that I observed and something that I wrote recently is that in a place like the United States, most of the people, half of the people in the country say to pick a round number, they'd have balance sheets that are so thin and fragile they can't, you know, survive the lockdowns and the pandemic economy. And they certainly don't own any of the assets that are experiencing this appreciation. This is just a personal view, and I don't think we have good models of how the economy works dynamically with these kinds of production functions and cost structures. We can measure unmeasured stuff like consumer surplus, and it's very important and Erik and his colleagues have done the best work on that. And we're about to learn about it, but that's different than an understanding, you know, of how an economy dynamically works both in functioning and distribution with this kind of structure.

So here's a picture. A very smart guy at an asset management company, he's came from Bloomberg. I just saw him. He basically divided the S&P 500 into deciles. I think he was just counting companies, not market value. And then look at them and then the top decile was the ones that have the highest value, market value per person, per employee in the company and highest intangible market assets, the market value of intangible assets. I mean, basically that you do that by subtraction. You take the market value and subtract assets and liabilities, if there are any, and this is the picture you get, these outsized bits of economic performance. This is year to date. So it's the pandemic economy are all, you know, or tend to be, or highly correlated with this sector that has high market value of intangible assets per employee.

And as you go down, then you get into other sectors that are lower, that generate value in normal times with some combination of labor and tangible assets, mostly. I mean, everything is a hybrid, but the mix differs. And this is really interesting. I mean, basically it's an example of what I asserted before, only it's extreme. And it's an example of divergence in the markets induced in part by the pandemic economy. And basically the value creation is disconnecting from employment in the economy.

The other thing that people talk about a lot, now I'll have to do this very quickly or I'll just run out of time, is, there's a lot of writing about how can the stock market be either flat or up, the NASDAQ's up 25%, the S&P is as of the recent correction about flat, many economies are in such crappy shape. And the answer is that the stock market is, you're only puzzled by it if you look at one index, right? If you look underneath, you'll see the acceleration in the digital economy, and you'll see the struggling in other parts of the economy like airlines, which has a Dow Jones index that looks approximately like this. They were at some point down 60%, they're probably down 40% now, and they will struggle to recover. So, so the stock market is I think, an underutilized measurement resource. And if you get underneath the single index, then the puzzles go away and it's a reasonably decent reflection or mirror of a part of the economy.

Now, I admit if you went to the set of companies and people who work for them, and they are not publicly traded companies, I think things would probably look a fair amount worse for reasons that you understand having to do with the way that economy's put together.

Some of you seen this before, I'm just going to leave it with you. A study that I found interesting done by two faculty members at University of Chicago asked how many jobs can be done at home. And the answer was for the American economy, 33%. And then this is across sectors. So education, 83%; professional scientific, 80%; central finance, 76%; tech companies, 72%. And that's a pretty broad category. You go down to the bottom, accommodation and food service employees, 16.7 million people, of whom 4% can work at home. The rest are unemployed or were, and you can see that this kind of divergence matters. I wanna, maybe there's a little advertising along, but there have been some serious attempts at capturing important dimensions of digital impact. And one of them was the first report that the Luohan Academy did called Digital Technology and Inclusive Growth.

And what it did for those of you who haven't had a chance to read it, is it documented for the Chinese economy, mainly, the inclusive results of the rapid growth of e-commerce, mobile payments and fintech. And it's only part of the digital footprint even, you know, in China and certainly globally, but it's an important part of the footprint. And we've seen more muted versions of that in other places, but we saw big accelerations in places that fallen...virtually everybody's fallen behind China in some of these dimensions because of the pandemic economy, because we basically were forced to change.

And so I don't mean to say digital technology is good, but let me state a hypothesis based on this scanning exercise I've been doing for the last couple of years. Many of the positive effects of digital technology, regardless of which sphere they occur in, look to me to fall in this inclusiveness category. I'm not saying that that means, you know, I'm not dismissing its long run potential to affect productivity and other things, but it does seem to have huge potential in the dimension that Long and his colleagues wrote about in this report. That said, I mean, I just wrote down at the bottom that does not dismiss the real challenges that we know exist in work skills, robotics, automation, and other things. I really just wanted to say a word or two about resilience. It seems to me that one of the things that pandemic economies did is bringing resilience on the list of potential sort of things to worry about kind of up on the list, um, elevated its weight or importance, and digital technologies, I think, most of us would agree have been from many countries, the key element of resilience in the pandemic economy, because they enabled parts of the economy to function in the face of extreme mobility declines.

Some, however, systems fail badly. And these, I think, deserve attention. I'm not sure this is exactly a measurement problem, but it would be nice under the heading of resilience to ask, have we really done a good job? So for example, we had in many American states, virtually all, in fact, huge spikes in unemployment claims and in a significant fraction of those cases, the systems, the IT systems simply crashed. And so that's produced anxieties, delays and other things that have potential real costs in terms of people's ability to survive or company's ability to survive. And this is a measurable sub problem, if I can put it that way, and it's actually pretty easily fixed with cloud computing.

The other thing that I wanted to mention under the heading of resilience. Resilience has many, many dimensions, and we could spend a couple hours on it, but it's very striking that many, at least some of the Asian economies, including China, have dealt quite effectively with the pandemic economy, with the virus using, in part, digital tools. And what they basically dealt with is something that my friend Mohamed El-Erian calls "human counter party risk". Everybody knows what counter party risk is. You know, when the counter party is in financial transactions that don't trust each other, this transaction stop happening, it can bring a financial system down. You know, as we know from various episodes historically, including the GFC where it almost... It could have happened again. This is the analog, right? So, you know, absent, uh, some mechanism for sort of differentially affecting people's mobility depending on their infection risk. You basically have a situation in which everybody's equally risky and that produces, because of risk aversion, significant headwinds to the development of business.

Asian economies have, in China in particular, I think has done a good job of dealing with that. In the West, our concerns about privacy, data security, distrust of business and government caused us to under-utilize these tools. So I cite these examples because I think they're interesting. Finally, this is my last slide. I will say, you know, indifferent to Bob Gordon, um, intermediate, this is a point I made a year ago, but I'm going to repeat it. GDP and its flip side income is really important, and nobody is proposing this sort of throw-it-out of a multidimensional well-being measurement system or system of trying track our progress. So the question is, if you accept that there's some important work to do to understand its effect on various dimensions of well-being, does digital technology matter for GDP growth, for incomes, for productivity?

I don't want to... Just leave out free services for now. We know GDP doesn't measure consumer surplus... Let me leave it out so Erik will bring it back in, but just leave it out for now. As we know, consumer surplus isn't in there along with a whole lot of other services that get delivered that don't have markets and prices. And finally we get to the point where the digital technologies, if they're going to affect GDP growth, are probably in economic terms. They're going to have to be intermediate goods, meaning they're not part of final output, they're things that are needed to produce final output. And if they're powerful enough, then they'll make it, they'll increase productivity. And I think, there are gray areas. I mean, as soon as I said that, about six months ago, or somebody said that a self driving car or a robot that cleans your house, you know, is that a consumer good, a part of final output, or? the answer is I don't know.

And so I think it's fair and important to ask if digital technologies are... If our measurement systems are gonna be able to capture the impact of digital technologies on productivity, on labor targeted activity in particular and on growth, and given the recent trend in productivity globally. If the answer to that is ever going to be yes, it looks like it'll have to be in the future because we're just not there yet. Um, and so I don't think that's a reason for sort of dismissing the proposition that digital technologies are going to have an enormous impact in a wide range of fields. But I think the jury is out on the question and maybe the jury isn't. Bob Gordon will tell us, it's already decided. I think it's...the jury is out on this question of whether digital technology is actually gonna make a big difference to growth.

 

For more information, please visit Luohan Academy's youtube channel: Luohan Academy

    Event materials

      0 comments

      to leave a comment

      loading...

      00:00:0000:00:00