Stories of American renewal

James and Deborah Fallows’ book Our Towns, and James’ Atlantic article on the same subject, are getting a bunch of attention. I haven’t read the book but recommend the article. Fallows writes:

America is becoming more like itself again. More Americans are trying to make it so, in more places, than most Americans are aware. Even as the country is becoming worse in obvious ways—angrier, more divided, less able to do the basic business of governing itself—it is becoming distinctly better on a range of other indicators that are harder to perceive. The pattern these efforts create also remains hidden. Americans don’t realize how fast the country is moving toward becoming a better version of itself.

The basis for this is (largely) years spent traveling and reporting across the country. Which raises the question:

Were we mistaking anecdotes and episodes for provable trends? This is the occupational hazard of journalism, and everyone in the business struggles toward the right balance of observation and data. But the logic of reporting is that something additional comes from traveling, asking, listening, seeing. This is particularly true in detecting a sense of changed course. A political movement, a new technological or business possibility—I have learned through the decades that enthusiasm in any of these realms does not guarantee world-changing success, but it’s an important marker… And enthusiasm is what we have seen.

It’s easy to finish the article and find yourself wondering if you believe it. Could the Fallows’ story of American renewal possibly be right? Since reading the article I’ve come to think it is best paired with another recent one, by Walter Russell Mead at Foreign Affairs. He is also taking stock of the current moment in America, but where the Fallows report from the ground, Mead takes the 30,000 foot view. He begins by listing America’s difficulties during the industrial revolution. Mead continues:

The United States is passing through something similar today. The information revolution is disrupting the country’s social and economic order as profoundly as the Industrial Revolution did. The ideologies and policies that fit American society a generation ago are becoming steadily less applicable to the problems it faces today. The United States’ political parties and most of its political leaders lack the vision and ideas that could solve its most urgent problems. Intellectual and policy elites, for the most part, are too wedded to paradigms that no longer work, but the populists who seek to replace them don’t have real answers, either. It is, in many ways, a stressful and anxious time to be alive. And that anxiety has prompted a pervasive sense of despair about American democracy—a fear that it has reached a point of dysfunction and decay from which it will never recover.

The effects of rapid change are often unwelcome, but the process of transformation is one of growth and development, not of decline and fall. Indeed, the ability to cope with change remains one of the United States’ greatest sources of strength. In the nineteenth century, people often compared the United States unfavorably with the orderly Prussian-led German empire. Today, the contrast often drawn is with
China’s efficient modernization. Yet there is resilience and flexibility in the creative disorder of a free society. There are reasons to believe that, once again, the United States can find a path to an open and humane society that capitalizes on the riches that the new economy will produce.

Mead’s is the sort of article so sweeping as to inevitably raise some objections. But I have found myself thinking of his and the Fallows’ arguments as stronger when considered as part of a whole. The optimistic case for America need not hinge on whether American towns are on the way up or down — whether or not they are, on net, undergoing something that can be described as renewal. The optimistic case simply requires that they possess that potential. We’re going through a period of tremendous disruption, but if the Fallows are right then Mead is: we may have what it takes to make it through, stronger and more prosperous.

Two things to read on paywalls

Paywalls and the business model of journalism were one of the early focuses of this blog, but I don’t write about them much anymore. That’s odd, given that I still think about them quite a lot and that they’re part of my current work.

I don’t have anything to say about them right now, except these two pieces are each worth a read:

Megan McArdle on “A farewell to free journalism” at The Washington Post:

Critics of the “mainstream media” (or if you prefer, the “lamestream media”) are fond of saying that we’re going to be put out of business by competition from “new media” upstarts. Indeed, as a young blogger, I might even have made a few such pronouncements. And I and those critics were wrong. Traditional media can survive competition for readers just fine. It’s competition for advertisers that’s killing us.

John Micklethwait, editor of Bloomberg News:

But is journalism really in such a parlous state? Look closer. News is an industry in transition, not in decline. It is reemerging as something more digital, more personalized, more automated, more paid for—and (eventually) less fake. In many ways history is repeating itself, with the main surprise being the survival of so many established names. And good journalism still does have the power to change lives.

Heterogeneous treatment effects

I wrote two posts in the past couple of months about when we can be confident that social science results apply in specific cases (or not). Specifically, the posts were about when we can trust average causal (treatment) effects, given that the average effect might mask any number of hidden interaction variables or, put another way, differences in treatment effect across group members.

I’m pleased with how those two posts came out, but they both assumed that average treatment effects were all we had. I focused on practical questions you might ask to decide whether to trust the social science, not on how the social science itself could answer the question.

I saw a good presentation recently on machine learning and heterogeneous treatment effects, so I thought I should post something mentioning that social scientists do often try to go beyond just main effects to consider how the effect might differ for different parts of the population. A bit more on heterogeneous treatment effects is here, and a good paper on how machine learning can help estimate them is here.

Update: And one more good link, from Microsoft Research.

Reporting on ideas

Here’s James Fallows, of The Atlantic, on his reporting style, specifically after moving to China:

The best way I know is just to have a big list of the things you’re genuinely curious about, ask people to answer what they view of them, and just sort of triangulate, and see the range of opinions you get — what people are agreeing on, what the outliers views on, and just keep trying those hypotheses… I would never call journalism science, because it’s craft… [but] there’s a quasi-scientific method of hypothesis testing. You say ‘OK this is what I’ve heard, does this sound right to you?’ … and you share that work with the reader.

Here’s economist Bryan Caplan on how he does interdisciplinary research when writing a book:

When I’m going through and reading papers, oftentimes I’ll find that there’s something I don’t understand very well or something that seems questionable. Usually, as long as the authors are living, I do try to actually reach out to them and get clarifications.

The next thing is when I’ve got what I think is a good, solid draft of a book. That’s where I enlist my RA and say, “Get me all the emails of all the living people I’ve cited in the book so far,” and then email all of them with two offers.

One of them is an offer to show them the entire manuscript. The other one …  is to say, “Or, if you’re busy, then I could just tell you the exact pages where I discuss your works, so at least you can tell me whether I’m accurately summarizing your work or not.”

For me, what I do is so interdisciplinary so I’m always worried about this autodidact’s curse, where you’ve read a ton of stuff but you still haven’t actually talked to anyone who knows what’s going on. This is one of the things that I try to do to deal with especially the wisdom of a field. Oftentimes there’s wisdom in a field, where it’s known to people who have thought about it for a long time, but they don’t write it down.

Of course, that’s very hard for the autodidact to find out. “What is the wisdom in your field that you don’t write down?” This is where I try to reach out to people. Generally, I would say I get about a 15 percent response rate for the people saying they’ll at least read something, so I feel like it does give me some good quality control.

This sort of “reporting” is the kind I get excited about and most enjoy doing, and I see it as quite in line with Fallows’ method. Opinions probably differ on whether this sort of process is “journalism”, and it probably depends on a bunch of different factors. But to me, this sort of reporting on ideas is where all the fun is. And when it’s done well, sometimes you can even get a bit beyond where a field of inquiry is, by drawing on lots of different disciplines, spelling out that wisdom that hasn’t been written down, and answering questions that experts know something about but haven’t yet answered definitively in their published studies.

When critical thinking helps

The most politically engaged people are also the most partisan, and in many cases they’re also the most resistant to revising their views. They even — if I’m recalling this Ezra Klein show episode correctly — read articles by the opposition more closely so as to better pick apart their opponents.

In other words, the fact that you care about politics and are highly engaged isn’t evidence that you’re objective, and might in fact indicate the opposite.

Listening to all of this, you might think critical thinking is just a facade, that it’s purely a way to justify the conclusions we’ve already reached. As I’ve written before, that’s far too cynical. And here’s some new evidence that’s more optimistic, via Nieman Lab:

These studies also suggested that two related forms of thinking may protect against belief in fake news: The first, actively open-minded thinking, involves the search for alternative explanations and the use of evidence to revise beliefs. The second, analytic thinking, involves deliberate thought processes that consume memory resources.

Critical thinking and open-mindedness really are helpful! And this is hardly the only paper to find that result.

For more in this vein, I wrote a few months back about decision-making for HBR, drawing on research on how to think and decide rationally.

A framework for thinking about market power and rents

I’ve been thinking a lot about market power, rents, “superstar” firms, etc. — here’s my reading list — and one important thing to keep in mind is that firms can make excess profits for different reasons. And these different sources of rents can be more or less harmful.

But how should we think about the many different sources of rents? In my view, combining one theoretical distinction with an empirical one leads to a helpful framework.

In terms of theory, rents can be earned because of an industry’s structure, or because of differences in firm capabilities. That’s an imperfect distinction, of course. Michael Porter would note that successful firms tailor their strategies (and therefore their capabilities) to their industry, and their position in it. Nonetheless, this distinction is a helpful starting point.

Industry structure-based rents are a focus of Industrial Organization, a subfield of economics, and include barriers to entry, economies of scale, and the like. The study of firm capabilities happens more often within the strategy and management literatures (and in particular the strategy paradigm known as the resource-based view). The idea is that some firms are better managed than others, or have unique or hard to copy assets or capabilities, and these can allow them to earn excess profits.

Then there’s a simple empirical observation about the rise of rents and market power over the past 20 years: a big part of it has to do with digital technology. Of course, not all of it does, so it’s worth distinguishing between the role of digital tech and other sources of rents.

And so that leads to four buckets for categorizing potential sources of market power:

market power sources

Analog capabilities and assets includes good management, a strong brand, etc.

Digital capabilities and assets includes data, and expertise at building software or using analytics, for example.

Analog industry structure includes regulation, non-technology-related economies of scale, etc.

Digital industry structure includes economies of scale from IT, both on the demand and supply sides.

Some things don’t fit so neatly. IP is an asset, but it can be related to digital or not, and the importance of IP depends arguably on industry. Lobbying is a capability, but it’s more important in particularly regulated industries.

This isn’t the only way to think about this, and I’m certainly not implying that all four boxes are equally important! Nonetheless, I think this is a helpful way to bucket the various sources of market power. If some companies are temporarily making excess profits because they’re better adopters of technology — digital+firm — that’s not necessarily a problem. On the other hand, if companies are writing regulations in their favor — analog+industry — that’s more worrisome. If it’s mostly about digital scale economies — digital+industry — that may be somewhere in between.

One last point: research suggests that analog intangibles, like management, are complements to digital tech. Companies with good management are better at IT and vice versa. So the two quadrants on the right side reinforce each other in interesting ways.

Even if this framework doesn’t appeal to you, it’s important to consider the why behind market power. Different sources may be more or less problematic, and may need to be addressed in fundamentally different ways.