‘This is a story of people sleepwalking into power.’
‘This is a story of people sleepwalking into power.’
Part One of a conversation with Henry Farrell and Abraham Newman, authors of ‘Underground Empire: How America Weaponized the World Economy’
Abraham Newman is a professor in the School of Foreign Service and Government Department at Georgetown University. Henry Farrell is the SNF Agora Professor of International Affairs at the Johns Hopkins School of Advanced International Studies. Together they’re the authors of Underground Empire: How America Weaponized the World Economy.1
A few weeks ago I spoke with Henry and Abraham about what the “underground empire” is, how history is often shaped by inertia as much as by intentionality, what it was like to have the Trump administration take an article Henry and Abraham had published as a warning and instead use it as a policy “playbook,” and how science fiction novels influenced their writing.
This is the first part of our conversation; the second part is available here. The excerpts below have been condensed significantly and edited for clarity.
ADAM: What is the underground empire?
HENRY: The underground empire is this system of more or less disregarded parts of the global infrastructure, all of the networks that hold the globalized economy of the world together. We haven’t paid much attention to these until the very recent past because they just seem to be, fundamentally, pretty boring.
If you look at, for example, the ways that the internet works—all of these technical details of fiber-optic cables. If you look at the financial system, it depends on the dollar clearing system and the SWIFT system of financial messaging. All of these things seemed, until relatively recently, to be really dull, to be really mundane. You were very happy that they were there because they made your life easier, but they didn’t seem to be particularly important.
Now, to quote from Master and Commander, networks have become battlefields. In September 2023 we’re now in a world where all of these boring-seeming systems have become tools of power coercion. This is the story that we try to tell in the book.
One of the points that you make throughout the book is how haphazard and unintentional and accidental the emergence of this underground empire has been. After 9/11, you write, the U.S. government “stumbled onto the political power hidden amid the plumbing of this new global economy.” At one point, you talk about the U.S. “sleepwalking” into this empire. [You argue that] the “dollar became a global currency, without anyone really planning it.” Can you talk about the role of chance and randomness in all of this?
ABRAHAM: There’s two levels of that. On the one hand, firms made decisions that built this structure, that then opened up the power that we talk about. A lot of those decisions are based in entrepreneurs who are trying to make money.
We tell the story of Walter Wriston. He was, among other things, the head of Citicorp, and he came up with this innovation, certificates of deposit. At that time, he was just trying to figure out a way for people to make more money off interest. But that led to this dollar clearing system, eventually, that then gives the government immense power.
[Consider] the concentration of cloud servers near Washington, DC, [and] Amazon’s decision to build its cloud “hub” there. All of these things are not laid out in some master plan by the government or by the firms. It’s really about people making decisions often that accrue on each other.
We use the [phrase] “path dependency.” There were [already] cables laid in Ashburn[, Virginia, where the U.S. government built some of the first infrastructure of what became the internet], and so then they built these cloud hubs there. So on the one side the firms are making these decisions, which are incremental and build on each other, and also accidental.
And on the other side, on the government side, what our research shows is that these are basically civil servants, bureaucrats, people. They’re struggling with wicked problems—terrorism, the proliferation of nuclear weapons, border conflicts—and they’re trying to figure out, what can they do to solve these problems?
We really think the story is a story of people sleepwalking into this power, rather than really out there trying to get it and define it and create some orchestrated system.
HENRY: It’s very easy for human beings, given the way that our cognitive architecture is set up, to attribute a ton of intentionality to these systems, to think that, in effect, there’s some kind of grand, secret conspiracy behind the scenes, which is orchestrating this.
Our book is exactly about how this is not the case, how this huge and important system emerged not through any grand master plan, but instead through a series of accidental discoveries. People tried to capitalize on these discoveries and [built] the system in bits and pieces, never quite realizing what the whole was going to amount to.
You mentioned the concept of path dependence. Can you explain what that is and how it relates to this underground empire?
ABRAHAM: Path dependence is a concept in economics. It comes out of this idea that once people invest in a certain system, then they keep investing more [in it].
The quintessential example is the QWERTY keyboard. The QWERTY keyboard was invented because when [people] were using the typewriters, they didn’t want [keys] to be close together so that they would get stuck. What that meant was that the most common pairs of letters were put far away from each other.
It’s a really inefficient keyboard. But we all learned how to type on this keyboard, and to retrain everybody would cost a lot of money, and also the companies—they don’t want to change all their keyboards. It’s a very banal example of how we make a decision that’s actually inefficient today, but it stays with us over time because of these investments.
That happens in politics all the time. We create a system. That system in [a] crisis is all we have. So we build on that, even though it’s maybe not the perfect solution to the problem at the time.
There might be actors in the world—let’s say China. They would prefer to have a different global network for finance and for clearing trades globally. But so many other people have already invested in the dollar clearing system that everybody keeps using it because of that path dependent effect.
Something you talk about a lot in the book is the power of maps, not just to reflect the world, but to actually shape it. You argue that America “succeeded in mapping the world and hence in remaking it.” Tell us about the NSA’s STORMBREW map, and how it helped turn “the globalized economy into a territory that the United States could dominate,” as you put it.
HENRY: The STORMBREW map is a map that was part of the [Edward] Snowden revelations. It shows an outline of the continental United States with a small number of points, which are given code names taken from various skiing resorts.2
These points that were labeled in the map were actually key points in the global communications infrastructure. They were places where telecommunications cables landed on shore, or where there were switches where one major telecommunications system talked to another. And the United States figured out that these were places where it could tap into fiber-optic cables.
The STORMBREW map is a map of one particular part of the system. It isn’t the only map; there are many other maps. But once you have these maps, once you are able to identify these choke points, you’re able to turn what seems like an incredibly vast and complicated system—in which everybody is yelling at everybody else at once—into something that is clear, that is crisp, and that is actionable.
Once you have these maps, you can turn this very, very complicated global economy into something that you can act on.
I was struck by how much of the underground empire is still physical, even though the way that most of us interact with it is through mediums that are electronic and wireless and invisible. Server farms, fiber-optic cables, undersea infrastructure, semiconductors—what [are] the implications of so many of these sources of power residing in the real, tangible, physical world?
ABRAHAM: There’s both the physical of the network structures—you mentioned the cables, the semiconductors—but there’s also the locations. Organizations like Amazon, which is located in the United States, primarily—that gives a lot of legal power to the United States to grab it.
We tell the story of the SWIFT organization [which handles international financial transactions]. It’s based in Belgium. It’s a company that basically is an encrypted, private post office for banks. If you make a global transaction that’s over $10,000, the SWIFT system is basically sending the messages back and forth.
For a long time, [SWIFT] kept a server system located in the United States to mirror all of its data. It had one in Europe, but also one in the United States. And when the United States after 9/11 wanted to figure out how are the terrorists getting their money, they used the fact that there was this physical location of this data in the United States to get jurisdiction over it.
We often think of the cloud or data as something that just floats in the air, but it all rests on these very tangible structures that then provide legal, but also physical, jurisdictions for the government.
And one of the things we’re seeing also right now is governments trying to think, Can we restructure these infrastructures in a way that would get us more protection from these conflicting demands? One of the problems is that often the investments are hard to restructure because they’re actually physical.
We tell the story of the precursor to the cloud centers in Ashburn, Virginia. It was a company called Equinix. It built all of these fiber-optic cable connections. And one of the things that the head of it told us was, Once we dug all that cable, nobody’s going to shift, nobody’s going to change, because it [would be] so expensive.
HENRY: In addition to the kinds of physical infrastructures that Equinix and other companies like them represent, there also is what you might call [a] “geography of intangibles.” Over the last couple of decades, a lot of the physical infrastructure has shifted, so that the United States is no longer as central to, for example, global information and telecommunications infrastructure as it was in the early 2000s.
But what has not changed nearly so much is the location of intellectual property, which is about as intangible as you can possibly get. And so here we see this very interesting intersection between intangible intellectual property, and the tangible fact that many of the possessors of this intellectual property are still located on U.S. soil and U.S. territory.
Here we’re talking about semiconductor designers such as Qualcomm, [and] the various companies such as Cadence that allow for design to actually happen in the first place. And the United States has been able to use its control of these intangibles to really try to reshape the way in which semiconductor manufacturing works. And this is the basis of its efforts to deny China access to the most recent and most advanced semiconductors for AI.
This is the first of a two-part conversation. You can read the second part here.
https://us.macmillan.com/books/9781250840554/undergroundempire; https://www.ft.com/content/d00426ad-19d9-4e0f-be75-93c6b6a8cc64
Some of the names of points were Breckenridge, Whistler, Tahoe, and Copper Mountain: https://henryfarrell.net/wp/wp-content/uploads/2022/10/Stormbrew-map.pdf