The arc of the social universe is long, however it bends

Technology can be used for good or ill, but over the course of human history it’s clearly been a source of major progress. (If you’re not convinced, Robert Gordon’s book is one of my favorites to make the case.) But even if technology ultimately contributes to progress, it can cause a lot of harm along the way. I’ve just happened across a couple of reminders of that lately, and so wanted to put them (along with some others) together.

John Lanchester writing about the transition to agriculture in The New Yorker:

So why did our ancestors switch from this complex web of food supplies to the concentrated production of single crops? We don’t know, although  Scott speculates that climatic stress may have been involved. Two things, however, are clear. The first is that, for thousands of years, the agricultural revolution was, for most of the people living through it, a disaster. The fossil record shows that life for agriculturalists was harder than it had been for hunter-gatherers…Jared Diamond called the Neolithic Revolution ‘the worst mistake in human history.’ The startling thing about this claim is that, among historians of the era, it isn’t very controversial.

And here is Leandro Prados de la Escosura, writing in the Cambridge History of Captialism Volume II. The chapter is on capitalism and human welfare:

Trends in human development do not match closely those observed in real GDP per head. More specifically, phases of economic globalization have a dramatic impact on per capita income growth but not on the progress of human development. A counterintuitive lack of association is observed between human development and per capita income prior to World War I. Although the initial large-scale progress in health can be traced back to the late nineteenth century, with the diffusion of the germ theory of disease, and primary education experienced a significant advance in the era of liberal capitalism, the progress in human development dimensions fell short on the economic advancement resulting from globalization and industrialization. The negative impact of urbanization on life expectancy and the lack of public policies on education and health may account for human development’s slower progress in the late nineteenth century. More significantly, while real GDP per head stagnated or declined during the globalization backlash of the interwar years, human development progressed steadily. Health and education practices became increasingly globalized during the economic backlash of the period 1914 to 1950. Could a delayed impact of economic globalization on human development be, perhaps, hypothesized? Since 1950, advancement in human development has been hand in hand with growth in the world economy, although at a lower pace during the Golden Age (1950-1973) and, again, since 2000.

The bit on urbanization echoes Robert Gordon’s account, in The Rise and Fall of American Growth, that America’s shift toward cities in the 19th century “brought with it a host of public health problems.” From our vantage today, urbanization looks like a good and necessary thing. But for many if not most people, things got worse before they got better.

The last thing I’ll add here is James Bessen’s book Learning by Doing, which looks at the wages of weavers during and after the Industrial Revolution. His thesis:

Workers can benefit by acquiring the knowledge and skills necessary to implement rapidly evolving technologies; unfortunately, this can take years, even decades.

These examples are worth bearing in mind when we think about human progress in relation to things like globalization and technology today. Arguably, many ‘elites’ are waking up to the fact that they ignored the possibility that the benefits from economic changes that seemed likely to be net positive (trade with China, the adoption of IT) would cause dislocations measured not in months but in decades. But the historical record suggests that these painful lags are the rule rather than the exception. The discovery of agriculture was net positive, as were urbanization, the industrial revolution, and global trade. But every one of them came with massive suffering, experienced over long periods of time. The answer isn’t to be complacent and wait it out. It’s to undertake the necessary actions to lessen this suffering. To make the investments in public health that improved urban life; to erase the lag between the financial benefits of globalization and the benefits to overall well-being.

Nowhere is this more important today than with technology. Overall, technology has been a force for good. I think it can continue to be. But if we are not vigilant — if we do not pay attention to the harms that it causes, and do not prioritize investments to deal with them — then it will cause a lot of suffering along the way.

What is financial innovation?

Emphasis mine:

Whatever the cause or focus of financial innovation, it involved four basic forms, namely product, market, organization, and regulation. Product innovation was the design of a financial instrument to meet a specific need. Market innovation was the creation of a trading system through which these financial products could be bought and sold. Organizational innovation involved the grouping of separate financial activities together in such a way as to be greater than the sum of their parts. Regulatory innovation was a response to the need to reduce the risks involved and so increase the use made of the product, market, or organization.

That’s from the financial capitalism chapter in the Cambridge History of Capitalism, Volume II. The author is Ranald Michie. That last line is worth keeping in mind during the continued debates over how we regulate the financial system.

Nobel Prize-winning psychologist Dan Kahneman: AI will surpass humans in emotional intelligence

Some economists with whom I work hosted a conference on the economics of AI in Toronto last week, and while I couldn’t attend, some of the videos are online. And Brad Delong has provided a “rough transcript” of psychologist Dan Kahneman’s remarks, which include this bit about emotional intelligence:

[????] said yesterday that humans would always prefer emotional contact with with other humans. That strikes me as probably wrong. It is extremely easy to develop stimuli to which people will respond emotionally. A face that changes expressions, especially if it’s sort of baby-shaped, are cues that will make people feel very emotional. Robots will have these cues. Furthermore, it is already the case that AI reads faces better than people do, and can and undoubtedly will be able to predict emotions and their development far better than people can. I really can imagine that one of the major uses of robots will be taking care of the old. I can imagine that many old people will prefer to be taken care of by robots by friendly robots that have a name and that have a personality that is always pleasant. They will prefer that to being taken care of by their children.

Now I want to end on a story. A well-known novelist—I’m not sure he would he appreciate my giving his name—wrote me some time ago that he was planning a novel. The novel is about a love triangle between two humans and a robot. What he wanted to know is how would the robot be different from the individuals. I propose three main differences:

  1. One is obvious the robot will be much better at statistical reasoning and less enamored with stories and narratives than people.

  2. The robot would have much higher emotional intelligence.

  3. The robot would be wiser.

More from Kahneman at the link. This seems like an underrated possibility to me. Too often commentary on AI assumes that as machines take over analytical tasks, humans will focus on emotional ones. All this reminded me of a piece I wrote for HBR a couple years back that included this bit:

In his own research Gratch has explored how thinking machines might get the best of both worlds, eliciting humans’ trust while avoiding some of the pitfalls of anthropomorphism. In one study he had participants in two groups discuss their health with a digitally animated figure on a television screen (dubbed a “virtual human”). One group was told that people were controlling the avatar; the other group was told that the avatar was fully automated. Those in the latter group were willing to disclose more about their health and even displayed more sadness. “When they’re being talked to by a person, they fear being negatively judged,” Gratch says.

This is something we’d traditionally think of as interpersonal or emotional work, and yet people preferred machines because they didn’t come equipped with certain types of social/emotional reactions. That suggests to me that the bar is even lower than we think. Machines will get better at EQ, but they’ll also be able to respond differently than most people do, in ways that give them an edge. A machine that’s almost as good at listening and perceiving might still be preferable to a human in some cases if the machine also knows, for example, not to be judgmental. A machine that is nearly as good as a human at having a conversation might be preferable in some situations if it’s a more generous conversationalist and asks you more about yourself. AI may one day pass an emotional Turing test. But I expect we’ll start using them in social contexts well before then because of the other advantages they bring.

Larry Summers gives a great answer on competition, concentration, and market power

His podcast with Tyler Cowen is worth a listen for many reasons, but this was particularly in line with my own thinking:

COWEN: Regular economics: one hears increasingly these days that the higher concentration ratios in the American economy are an economically relevant fact. We all know those ratios are up somewhat.

But at the same time, consumers don’t, in an obvious way, seem to feel the burden of monopoly. If you look just at product choice and variety, productivity may be slow, the growth of manufacturing output appears to be quite steady. There’s not obviously a break in that series where all the monopolies restrict output.

We have all these different pieces of data: high-share values, high measured profits, very steady output behavior, a lot of product variety. How do you think about the issue of monopoly in the American economy right now? Is it significant or not?

SUMMERS: Most things I might be right or wrong, but I’m confident.


SUMMERS: This is one where I’m not certain. On the one hand, Tyler, higher concentration ratios, higher profit — with lower investment — manifest in lower interest rates. More monopoly power fits the story. On the other hand, take a thing like Apple Pay. Apple Pay means that Apple, which is already the largest company in America, is even larger.

So you could say that’s more monopoly and more market power. Or you could say there’s a whole financial industry it’s now getting competed with from outside the financial industry, and so it’s making things more competitive.

Both those views have merit. I think the second may have more merit than the first. Some of the rhetoric one hears recently, which leaves you with the impression that we’re seeing an era of a lot of new Standard Oils, seems to me to be quite overdone, with respect to the facts as I understand them.

On the other hand, are there combinations of healthcare systems in cities where we go from having some real competition to there being one dominant provider and networks to consumers’ benefits.

COWEN: Clearly, there’s problems there.

SUMMERS: There’s almost certainly some problems there. Are there difficult issues when information is central, as it would be with a Facebook or a Google? Yes, there are difficult issues: privacy, reliance on information, networks.

But are those best thought of through the prism of antitrust and monopoly? I’m not at all sure that that is the best way to think about them. After all, in some sense, the products in those two examples are given away to consumers for free. One of the most important issues for economists to figure out over the next couple of years is how to think about these trends.

I think there are some selective grounds for concern. It’s much more likely that antitrust has been insufficiently tough than it is that it’s been too tough over the last decade. But I also think one needs to be careful about the fact that being a successful business will tend to cause you to have more profits. We usually think of that as a good thing, not a bad thing.

A few things I think Summers gets right: 1) lack of certainty; 2) attention to ways in which business may be getting more competitive; 3) separating the tech platforms out from the broader story.

My best attempt to grapple with these trends is here; my reading list on the subject is here.

Blue Bottle, brands, and market power

Nestle has bought a majority stake in Blue Bottle, the fancy coffee chain. Why? As The New York Times reports:

The rapidly expanding niche accounts for 15 to 20 percent of coffee consumed in the United States, according to the Specialty Coffee Association. Perhaps more important than the sector’s rapid growth is that it commands higher prices and generates bigger profit margins.

So much of the talk of marker power, markups, and concentration has focused on technology and in particular the big tech giants. But what if Blue Bottle is the more representative example?

I don’t mean to knock the product when I say that Blue Bottle’s success is to a large degree about branding:

Although Blue Bottle is one of the most important players in the third-wave coffee sector, it has distinguished itself from its rivals in significant ways. It has spurned many of the hallmarks of high-end shops — barista competitions, lengthy travelogues about journeys to find the perfect small coffee farm — while emphasizing the aesthetics and experience of a well-prepared cup.

The strategy is a reflection of the company’s founder, a soft-spoken, classically trained musician who began roasting coffee as a hobby while on the road with traveling orchestras. Mr. Freeman’s approach has less of the rebellious rock ‘n’ roll attitude displayed by competitors like Stumptown, and more of the quiet calm that one might associate with a slowly but well-brewed cup of coffee.

You can argue over whether some of these differentiators are about product or brand, but the news reminded me of a paper I read recently: a 2016 review paper on branding by Bart J. Bronnenberg and Jean-Pierre Dubé. Some relevant bits:

In the latter part of the twentieth century, the degree of concentration in consumers goods industries grew at a much faster pace than other industries in the US (Caves and Porter, 1980). By the end of the twentieth century, most consumer goods industries were dominated by a small number of brands commanding most of the share of sales (Bronnenberg, Dhar, and Dubé, 2007). Most striking, many of the dominant consumer brands in 1923 were still the dominant brands in their respective categories in 1983 more than half a century later1 , although the findings are predominantly in food categories.2


In general, the sales in these CPG product categories are very concentrated, as summarized in Table 2. On average across categories and Scantrack markets, the top brand commands 46% of equivalent unit sales, which is consistent with the analysis of food categories in Bronnenberg, Dhar, and Dubé (2011). The average four-firm concentration ratio is 79%


The key consistent finding is the persistence in market shares and the advantages to early movers (even if not for the first entrant) that “survive” long-term.


The body of literature and the collection of empirical evidence supports the long-standing notion that established brands constitute important barriers to entry

To recap, branding appears linked to rising industry concentration. Branding should increase markups by the definition used in this recent and widely discussed paper if good branding requires higher up front costs relative to marginal ones. And it represents market power to the extent it deters new entrants. You could imagine digital technology amplifying the importance of branding, but also perhaps creating opportunities for new entrants like Blue Bottle as the ways that companies brand themselves change.

Needless to say, this is all speculative, and not, I think, in tension with other common explanations of market power, like lobbying, network effects, etc. But concentration + market power is a big, complicated trend, likely with many partial explanations. Perhaps the rise of branding is one.

Digital economics in one paragraph

Search costs are lower in digital environments, enlarging the potential scope and quality of search. Digital goods can be replicated at zero cost, meaning they are often non-rival. The role of geographic distance changes as the cost of transportation for digital goods and information is approximately zero. Digital technologies make it easy to track any one individual’s behavior. Last, digital verification can make it easier to verify the reputation and trustworthiness of any one individual, firm, or organization in the digital economy. Each of these cost changes draws on a different set of well-established economic models: Primarily search models, non-rival goods models, transportation cost models, price discrimination models, and reputation models.

From a review paper by Avi Goldfarb and Catherine Tucker, which discusses the empirical findings for each of the points above.

3 very different views of economic models

The first is a famous one, from Milton Friedman, via Justin Fox’s The Myth of the Rational Market:

[T]he relevant question to ask about the “assumptions of a theory is not whether they are descriptively “realistic,” for they never are, but whether they are sufficiently good approximations for the purpose at hand. And this question can be answered only be seeing whether the theory works, which means whether it yields sufficiently accurate predictions.

As Fox writes:

To head off the obvious objection that it was ridiculous to think regular folks reason according to complex statistical rules, Friedman and Savage argued that billiards players couldn’t write down the physics formulas that underlay their shot selections but nonetheless acted as if they did.

Here’s quite a different take that I just came up on recently, from J.W. Mason of the Roosevelt Institute:

It seems to me that Deirdre McCloskey was right: Economics is not the study of the economy. Economics is just what economists do. Economic theory is essentially a closed formal system; it’s a historical accident that there is some overlap between its technical vocabulary and the language used to describe concrete economic phenomena. Economics the discipline is to the economy, the sphere of social reality, as chess theory is to medieval history: The statement, say, that “queens are most effective when supported by strong bishops” might be reasonable in both domains, but studying its application in the one case will not help at all in applying it in in the other.

Finally, a third view, from Dani Rodrik’s excellent book Economics Rules:

I wrote this book to try to explain why economics sometimes gets it right and sometimes doesn’t. “Models” — the abstract, typically mathematical frameworks that economists use to make sense of the world — form the heart of the book. Models are both economics’ strength and its Achilles’ heel; they are also what makes economics a science — not a science like quantum physics or molecular biology, but a science nonetheless.

Rather than a single, specific model, economics encompasses a collection of models. The discipline advances by expanding its library of models and by improving the mapping between these models and the real world. The diversity of models in economics is the necessary counterpart to the flexibility of the social world. Different social settings require different models. Economists are unlikely ever to uncover universal, general-purpose models.

But, in part because economists take the natural sciences as their example, they have a tendency to misuse their models. They are prone to mistake a model for the model, relevant and applicable under all conditions. Economists must overcome this temptation. They have to select their models carefully as circumstances change, or as they turn their gaze from one setting to another. They need to learn how to shift among different models more fluidly.

Would better antitrust rein in the 1%?

The American economy has grown more concentrated in recent years — in most industries, the top few firms account for more revenue than they did 10 or 15 years ago. This phenomenon appears linked to the decline in the share of national income going to labor, as opposed to capital.

Would better antitrust help reverse that trend, resulting in higher wages and less inequality? That seems plausible, and my inclination is toward stronger antitrust. But the answer may depend on the sort of inequality we’re talking about.

It’s helpful, I think, to consider two types of inequality: the gap between the 1% and the rest, and the gap between educated professionals/the top 20% and the rest. Both gaps have grown, albeit for different reasons. The terrific rise of the 1% is largely an American phenomenon, whereas the other is more global.

Generally speaking, the international nature of the latter inequality has led economists to look beyond policy for explanations, to things like technology or globalization. The localized nature of the 1% inequality has led to an emphasis on U.S. policy decisions.

But the theme of a small number of top firms pulling away from the rest in terms of productivity and wages — not exactly the same as industry concentration necessarily, but related — appears to be international.

That suggests to me that the sort of inequality that would be addressed by tackling industry concentration is the widespread gap between educated professionals and the average worker, not the gap between the 1% and the rest.

This is all speculative, of course. It could be that concentration is a necessary but not a sufficient condition for the rise of the 1%, such that we don’t observe the same inequality around the world but that antitrust would address it. But it’s worth remembering what Alvaredo, Atkinson, Piketty and Saez concluded in their paper on the 1%:

The most obvious policy difference—between countries and over time—regards taxation.

That’s not the only cause they highlight. But the spectacular rise of the 1% is a mostly American phenomenon and so likely the result of something specifically American. The rise of incredibly powerful, productive superstar firms doesn’t fit the bill.

A reading list on market power, superstar firms, and inequality

My best attempt at an overview of the corporate concentration issue, from a recent piece:

The basic facts are these: Most industries in the U.S. have grown more concentrated, meaning the largest firms account for a larger share of revenue. At the same time, corporate profits have reached all-time highs, despite lackluster rates of business investment. And the number of new businesses being founded has declined; the number of new growth startups being founded has risen, yet these firms struggle to scale. The cause of these trends is not clear. Theories include the rise of IT and the network effects it creates, less-rigorous antitrust enforcement, and lobbying and excess regulation.

And a reading list (I’ll try and update it, and let me know what I’ve missed):

The Pro Market blog at the Stigler Center has been excellent on this. Here are a few examples:

Economists: “Totality of Evidence” Underscores Concentration Problem in the U.S.

The Rise of Market Power and the Decline of Labor’s Share

Is the Digital Economy Much Less Competitive Than We Think It Is?

So has the Washington Center for Equitable Growth. 

Market power in the U.S. economy today

Is declining competition causing slow U.S. business investment growth?

U.S. antitrust and competition policy amid the new merger wave

A communications oligopoly on steroids: Why antitrust enforcement and regulatory oversight in digital communications matter

New federal antitrust legislation recognizes U.S. workers are not only consumers

The New America Foundation’s Open Markets program is focused on this:

(update: New America and the Open Markets program have parted ways)

Amazon’s Antitrust Paradox

The Economist has done several great pieces:

The problem with profits

Too much of a good thing

A giant problem

The rise of the superstars

So has The Atlantic:

America’s Monopoly Problem

America’s Monopolies Are Holding Back the Economy

So has ProPublica:

The American Way

These Professors Make More Than a Thousand Bucks an Hour Peddling Mega-Mergers

This is a great piece from Fivethirtyeight on the state of startups:

The Next Amazon (Or Apple, Or GE) Is Probably Failing Right Now

Neil Irwin on winner-take-all at The Upshot:

The Amazon-Walmart Showdown That Explains the Modern Economy

Noah Smith at Bloomberg and on his blog:

Monopolies Are Worse Than We Thought

The Market Power Story

America’s Superstar Companies Are a Drag on Growth

Tyler Cowen on rising markups: “the whole story just doesn’t add up”:

The Rise of Market Power

Productivity and market power in general equilibrium

Intangible investment and monopoly profits

More on the rising markups paper (linked below):

Karl Smith

Robin Hanson

Arnold Kling

Larry Summers comments:

“I’m not certain”

Litan and Hathaway on decline of new business formation:

Declining Business Dynamism in the United States: A Look at States and Metros

Jason Furman was involved in two papers on this topic:

A Firm-Level Perspective on the Role of Rents in the Rise in Inequality


Vox’s Matt Yglesias and Ezra Klein did a podcast on this:

The Weeds

Justin Fox on winner-take-all industries and firm lifespans:

America’s Most Winner-Take-All Industry, Visualized

The Fall, Rise and Fall of Creative Destruction

Greg Ip

A Provocative Look at the Harm From Corporate Heft

A few of the academic papers that are good starting points in my opinion:

Are U.S. Industries Becoming More Concentrated?

The Fall of the Labor Share and the Rise of Superstar Firms

Declining Competition and Investment in the U.S.

The Rise of Market Power and the Macroeconomic Implications

An interesting anonymous contribution, via Tyler Cowen

Large U.S. firms have always commanded monopolistic rents–think Dupont, Bethlehem Steel, IBM, GM/Ford/Chrysler in their respective heydays. However, several developments have worked to dramatically change how those rents are shared. Before the 1980s shareholder revolution, monopolistic rents of dominant firms were more broadly shared–not just with rank and file employees but with the local community. (link)

We’ve done tons at HBR on this…

…firms are failing faster:

The Biology of Corporate Survival

The Scary Truth About Corporate Survival

…on startups being started but not scaling:

The U.S. Startup Economy Is in Both Better and Worse Shape than We Thought

…on superstar firms, and the pay inequality that comes with them:

Productivity Is Soaring at Top Firms and Sluggish Everywhere Else

A Study of 16 Countries Shows That the Most Productive Firms (and Their Employees) Are Pulling Away from Everyone Else

Research: The Rise of Superstar Firms Has Been Better for Investors than for Employees

Corporate Inequality is the Defining Fact of Business Today

Corporations in the Age of Inequality

Investing in the IT That Makes a Competitive Difference

…on digital firms pulling ahead:

The Most Digital Companies Are Leaving All the Rest Behind

What the Companies on the Right Side of the Digital Business Divide Have in Common

Managing Our Hub Economy

(see also several of the items in the superstar section above)

…an interview with Jason Furman:

Competition Is on the Decline, and That’s Fueling Inequality

…on mergers:

Mergers May Be Profitable, but Are They Good for the Economy?

…on data, AI, and antitrust:

How Pricing Bots Could Form Cartels and Make Things More Expensive

Should Antitrust Regulators Stop Companies from Collecting So Much Data?

…on lobbying and rent-seeking:

Lobbyists Are Behind the Rise in Corporate Profits

…on common ownership:

One Big Reason There’s So Little Competition Among U.S. Banks

Warren Buffett Is Betting the Airline Oligopoly Is Here to Stay

…on big companies paying more than small ones, but less so than they used to:

Big Companies Don’t Pay As Well As They Used To

…and trying to sum it all up:

Making Sense of Our Very Competitive, Super Monopolistic Economy


Will Wilkinson on dynamism and the welfare state

When I started writing about the link between entrepreneurship and the welfare state a couple of years ago, the research I was reporting on seemed to me to contradict a common belief: that economic dynamism was at odds with more expansive government.

But there are more and more signs that the prevailing wisdom is changing. Will Wilkinson had an excellent op-ed in The New York Times, notable in part because of his libertarian leanings. He writes:

Republicans need to recognize finally that secure property rights, openness to global trade and a relatively low regulatory burden are much more important than fiscal policy for innovation, job creation and rising standards of living…

Republicans make a critical mistake, both factual and strategic, when they cast the welfare state as the enemy of prosperity and freedom. Far and away the biggest items in the budgets of advanced countries are “welfare state” social programs, like Social Security, Medicare and Medicaid, which account for nearly half the federal budget…

dynamic, competitive markets drive growth and opportunity through “disruptive innovation,” which necessarily destabilizes industries and displaces workers. The lack of a robust safety net capable of keeping people from becoming collateral damage of routine creative destruction, such as the automation and offshoring of jobs or the closing of local plants, generates anxiety and hostility to the sort of economic policy that produces growth and employment. So not only are sound safety nets popular, but they also increase the public’s tolerance for the dislocations of a dynamic free-market economy.

Other signs of this shift include the Kauffman Foundation, which last year produced a policy digest arguing that expanding social insurance would boost entrepreneurship, and more recently Mark Zuckerberg’s commencement speech, which made this point without explicitly mentioning government.

More broadly, books like Concrete Economics and American Amnesia have argued that strong and aggressive U.S. government deserves credit for the nation’s history of dynamism. (Mariana Mazzucato was well ahead of this curve, as her book The Entrepreneurial State came out in 2013.)

Part of Wilkinson’s point, though, is to distinguish between the redistributive state and the regulatory state. It’s the latter, he argues, that more frequently impedes innovation.

My view is somewhat less pessimistic, and my bottom line in 2015 was:

Even the assumption that bureaucratic “red tape” holds back startups is less obvious than it sounds… What evidence we do have squarely challenges the intuition that it’s government that holds back startups.

But if Wilkinson is going to acknowledge the entrepreneurial benefits of the welfare state, liberals ought to at least consider the possibility that regulations do hamper innovation.

Putting entrepreneurship and innovation aside for a second, it seems clear that the net benefits of regulation vary considerably depending on which ones you’re talking about. The Clean Air Act seems to have had large positive effects. On the other hand, overzealous land use regulations that prohibit building have had large negative effects.

The same is likely true of regulation and entrepreneurship. Plenty of regulations probably aren’t a big impediment; some even help. But plenty of others probably do hold back innovation.

If we can agree about the merits of the welfare state — or at least agree that it’s not a major barrier to innovation and entrepreneurship — perhaps we can start talking about which regulations are worth keeping, and which are in need of reform.