Good morning, everyone.  My name is Chris Ford, and I’m a professor of international relations and strategic studies at Missouri State University and Visiting Fellow at Stanford’s Hoover Institution, and a veteran of many years of government service, most prominently at the Department of State.  Thank you for inviting me.

As I’m obliged to do as a former senior government official, I have to warn you that I can only offer you my personal views here today, and that they don’t necessarily represent those of anyone else – in the government or otherwise.  I hope that doesn’t make them any less interesting, for I am very pleased to be here. 

The concept I’ll be addressing today may already be familiar to some of you.  It’s the phenomenon of “weaponized interdependence” as described by Henry Farrell of Johns Hopkins and Abraham Newman of Georgetown, in a 2021 bookthey co-edited with Daniel Drezner of Tufts.

So what I’ll try to do this morning will be to cover four main points:

• First, I’ll outline that concept of “weaponized interdependence” as Farrell and Newman describe it. 

• Second, I’ll try to describe how useful I think their construct is, even when all the formal requirements they outline for it are not met.

• Third, I’ll offer some illustrations – based upon my own time in government in two different presidential administrations – that help demonstrate the utility of such thinking, not merely as an explanatory construct but also as a theoretical framework to help policymakers think through some of the practical challenges of statecraft that they face.

• Fourth, I’ll describe how I think such concepts in fact already seem to play a central role in the grand strategy of the People’s Republic of China (PRC) and its ruling Chinese Communist Party (CCP).

In covering these bases, I hope I can leave you with an appreciation for the importance of thinking about issues related to the “weaponizability” of interdependence, as U.S. and other national leaders consider policy options in today’s environment of strategic competition, and as scholars seek to understand the complexities of modern international relations.

That’s a lot of territory to cover, but let’s give it a shot.

I.          Weaponized Interdependence

So what do Farrell and Newman mean by “weaponized interdependence”?  In effect, it is a concept that stresses the practical power that can be created when complex networks of relationships take a very specific form.

But first, before I get too theoretical in that vein, let’s back up for a moment and put this in historical context.

It wasn’t really until the beginning of the First World War that modern industrial economies ever had to mobilize for all-out war, and this almost immediately meant that the Allies and the Central Powers also mobilized for economic warfare against each other.  And in this economic mobilization, as Nicholas Mulder has recounted in his book entitled The Economic Weapon, the centrality of London’s financial and commercial institutions in prewar global commerce gave British leaders a powerful tool.  Due to their control over London, they could to a great degree determine who in the rest of the world was able to benefit from – and who would be cut off from – those global networks. 

This was basically due to a network effect: the topography of the global trading and financial system as of the summer of 1914 was such that all – or at least most – European commercial and financial transactions and engagements ran in one way or another through London.  This gave the British power to monitor this traffic, in support of what was then an emergent discipline of economic intelligence, but it also gave them the ability to restrict flows for strategic purposes. 

This was, in a sense, a novel model then, but nearly a century later – in a different  context – it remained a potent one.  From the late 1990s, as recounted by my former administration colleague Juan Zarate in his 2013 book Treasury’s War, American leaders were able to take advantage of a similar degree of structural network dominance by leveraging the centrality of U.S. banks and other financial institutions in the early 21^st Century world as they developed Washington’s modern toolkit of financial sanctions. 

So the phenomenon of exploiting a dominant position is hardly new.  Farrell and Newman took the further step, however, of seeing this specifically as a network effect, and of describing it in theoretical terms. 

Such dominance doesn’t happen just anywhere, Farrell and Newman argued, but only where the particular structure of global relationships has acquired a pronounced “hub-and-spoke” character.  That is, there’s one huge central hub with innumerable direct connections to essentially all the other nodes in the system, and relatively few cross-wise connections between those other nodes.  That hub, in other words, dominates the network, and there are very few ways for other nodes to engage with each other that don’t in some way go through that hub. 

This also helps explain the persistence of such structures over time.  The advantages of joining to the hub are high, not only on account of whatever benefit the hub is able to provide on its own, but also because that’s usually the easiest way – the one with the fewest separate nodal “hops,” if you will – for any one given player to be able to engage with any other. Moreover, the costs of constructing any kind of alternative network structure are high, since doing that would require somehow persuading great numbers of players to link instead to an alternative hub.  But that alternative hub, by definition would not at first – and potentially not for a long time – provide them with anything like the connectivity benefits of connecting to the big existing central hub.  As a result, smaller players tend to get “locked in” by the structure of the network. 

But this isn’t just a topography that should be familiar to complexity scientists.  We also see versions of it in the human world around us all the time, especially in the internet age – in things such as social media (e.g., Twitter [“X”] and Facebook), online commerce (e.g., Amazon), and Internet search engines (e.g., Google).  In these arenas, there seem to be huge rewards to scale, which can create kind of a “virtuous circle” of success, in which the more connected a given node is, the more incentive there are for other nodes to connect to it, thus making that super-connected node even morevaluable to connect to, and so forth.  Hence, for instance, our modern high-tech “hyperscalers.” 

The key conceptual insight here is that networks such as global finance and the Internet are networks like this, and that if you happen to be in a position to regulate access to those hubs, you have a great deal of potential power, because the “exit costs” for other players may leave them no alternatives.  As Farrell and Newman put it, such an understanding of the potential implications of network architecture offers a “structural explanation of interdependence in which network topography generates enduring power imbalances among states.” The “topography of global networks,” structures the availability of coercion,” and when such coercive power becomes possible, they call it “weaponized interdependence.” 

In their analysis, if you do end up having a dominant position within a network like this, you can use it in at least two key ways.  First of all, even without actually coercing anyone, you can use such a position for informational advantage, in the sense that it can basically let you see everything that goes on in that network: you can see this traffic because it all more or less runs through you. 

The second potential use of a dominant position is coercive, and consists of using one’s position to exploit the “chokepoint effect” created by everyone else’s network engagement having to run through you.  If you have a chokepoint advantage, you can cut off those who displease you, and they can’t do too much about it.

II.        A Useful Construct

I think the Farrell and Newman construct is very useful.  It’s an elegant way to think about the problem, and it has the advantage of being not just consistent with Complexity Science, but actually corroborated by real-world experience with the modern hyperscalers made possible by Internet connectivity and by the practicalities of historical cases such as British economic warfare in the First World War and U.S. financial sanctions since the late 1990s.

To my eye, it also opens the door to some interesting applications – perhaps even more than Farrell and Newman originally expected.  In their initial telling, after all, the idea of “weaponized interdependence” was envisioned in fairly narrow terms: it emerged only when a pretty strict set of criteria were met in terms of a given network’s extreme  degree of centralization around a single dominant hub.  Without that kind of really hypertrophic dominance, they didn’t regard “weaponized interdependence” as existing, and they focused their own analysis only upon two cases: global financial networks and the Internet, in both of which the United States had an enormously strong “hub” position vis-à-vis essentially all other players.

But their analysis, I would say, actually has a much broader applicability.  Farrell and Newman stressed, for instance, that in order to capitalize upon the kind of asymmetric network topography they described – that is, in order to weaponize that interdependence – a country has to have sufficient “regulatory capacity” to do so.  This, they said, was what distinguished their two cases of U.S. financial dominance and U.S. Internet dominance from each other: Washington possessed an institutional infrastructure and the willingness to exploit its financial dominance through financial sanctions, but it had neither a lawful mechanism nor a willingness to start cutting countries it disliked off from the global Internet.

This point about institutional capacity, however, is important, for it suggests that “weaponized interdependence” probably can be usefully thought about as a question less of kind than of degree.  After all, neither regulatory capacity nor one’s willingness to exploit interdependence are going to be binary, all-or-nothing things; they clearly manifest by degrees.  Nor, truth be told, do networks always assemble into hub-and-spoke architectures either “completely” or “not at all.”  There, too, there likely some gradation. 

So while I do think Farrell and Newman’s idea of “weaponized interdependence” is useful in its “strong form” – with truly hypertrophic network topographies such as that Juan Zarate and his Treasury colleagues became so good at exploiting in the mid-2000s – it’s also got a lot of utility by degrees.  And this is where I think their scholarship can help academics and policymakers alike structure their thinking about the economic instrument of statecraft. 

Thinking of “weaponized interdependence” as a question of degree rather than of kind suggests the potential that actors may be able – as a matter of national policy – to intentionally both (1) create or accentuate advantages within any given network of relationships, and to (2) develop and hone the institutional capacities needed to exploit such asymmetries for strategic advantage, even as they seek to (3)  undermine their competitors’ efforts to do such things.  Rather than just taking advantage of whatever hub-and-spoke asymmetries just happen to have emerged as a result of market forces, in other words – which is basically what the British were doing in 1914 and the Americans in 2005 – network-based leverage may actually be a key competitive arena of power, in which states and other actors can maneuver against each other to deliberately and systematically create and exploit topographic effects.

III.       Interdependence and its Weaponization in Practice

Indeed, just based upon my own last few years in government, it seems clear that policymakers both in the United States and elsewhere have already been approaching national security and foreign policy challenges in some of these basic sorts of ways for some while – albeit not, of course, using the specific Farrell and Newman framing or terminology.  Having realized the power it has in financial networks, the U.S. Government has actually spent a good deal of time honing what Farrell and Newman would call its “regulatory capacity” to exploit network effects. 

When I was last in government – and long before we had heard the term “chokepoint effect” from Farrell and Newman – we spent a lot of time at the State Department working with our Commerce Department colleagues on national security export controls as part of our overall effort to reorient U.S. foreign and national security policy toward strategic competition with China.  In that work, we explicitly used what we called “chokepoint analysis” in order to identify areas where it might be possible to impose controls to slow Beijing’s acquisition of high technology items that could support the People’s Liberation Army under China’s strategy of “Military-Civil Fusion.”

In this work, it was clear that not every export control restriction would work in the ways we might like.  But it was also clear that some probably would. 

We opted, for instance,  not to restrict sales of lower-end semiconductors to China, in part because these were widely available from other national suppliers and our restricting American exports wouldn’t really affect Beijing’s access.  We also exercised restraint there because the U.S. companies who sold those items to China used the profits from these sales to support R&D that kept American firms at the cutting-edge of the industry in other respects.  So going after garden-variety chip exports didn’t sound like a good idea at the time.   

But we did move against U.S. exports to China where our analysis showed the United States to have a particular “chokepoint” advantage, and where China lacked alternatives.  One such “chokepoint” technology was in the sophisticated software tools used to design high-end semiconductors, a sub-field which the United States essentially monopolized, and which we duly restricted for China in 2020 through adjustments to something called the Foreign Direct Product Rule (FDPR).  Another “chokepoint” was Dutch extreme ultraviolent (EUV) lithography for semiconductor chip etching, a high-end capability that no one else – including American companies – could replace, which led us to begin an intense diplomatic effort to persuade the Netherlands to restrict sales to Chinese semiconductor companies such as Huawei. 

So you can see here some distinctly Farrellian and Newmanian themes of working to leverage interdependence, as it were, avant la lettre.  In effect, we were actively looking for technology supply chain relationships that were, you might say, loosely analogous to the kind of quasi-irreplaceable “hub” situations those two scholars described.

Furthermore, in terms of building what Farrell and Newman call U.S. “regulatory capacity,” we also moved to add more categories of U.S. export recipients to something called the Commerce Department “Entity List.”  Sometimes this was in order actually to restrict supply – e.g., to limit transfers into China’s technology supply chain where our analysis indicated a useful “chokepoint” existed – but it was also sometimes undertaken simply in order to get better data about what was moving in trade in the first place. 

By way of background on that point, you should be aware that formally speaking, being on the Entity List doesn’t necessarily mean you can’t purchase goods from the United States; it just means that an export license is required in order for Americans to sell you something.  In any given case, such a license might be approved, or it might be denied. Sometimes listing would occur on the basis of a policy decision to establish a “presumption of denial” for licenses to a particular entity, but not always.  Sometimes requiring licenses was largely simply an informational move, intended to help us gauge whether we had a technology control issue in the first place, since generally trade can occur without notice (and hence visibility) to the government unless licenses are required. 

Anyway, in Farrell-and-Newman terms, such postings to the Entity List could be seen as further developing U.S. regulatory capacity both to exploit trade flows for informational advantage and also, if needed, to impose “chokepoint” barriers for policy effect.  And some of this was fairly innovative within the U.S. system, too, as the Entity List process as first established in 1997 was focused only on entities involved in WMD proliferation.  We expanded and routinized its employment, however, for “activities contrary to U.S. national security and/or foreign policy interests,” making it a much more broadly applicable tool.

You can also see dynamics that Farrell and Newman would find familiar in the efforts our sanctions targets made to get out of the network-topographic pickle we tried to keep them in.  Most obviously, after the United States pulled out of the Joint Comprehensive Plan of Action (JCPOA) arrangement with Iran in 2018, the Iranians and several European governments tried to establish a mechanism for trade between Europe and Iran that was not subject to restriction by U.S. sanctions.  This “Instrument in Support of Trade Exchanges” (INSTEX) system aimed to create an alternative commercial mechanism based around of non-dollar-denominated, barter-based transactions that was independent of the U.S.-dominated hub-and-spoke global financial system. 

As it turned out, however, the topography of that U.S.-dominated financial network was so compelling that it was very hard to find European companies willing to participate.  Even though their specific transactions with Iran might not  themselves involve U.S. dollars or pathways through U.S. banks, European firms could only truly immunize themselves against potential U.S. sanctions for trading with Iran by entirely severing all their ties to all U.S. financial networks or anyone who used them – and this was something that no sane European company was willing to do.  Accordingly, as Iran bristled at the Europeans’ inability to offer more than merely humanitarian goods, to which the Americans were expressly willing to permit, INSTEX was eventually disbanded after having processed only one single transaction. 

Henry Farrell and Abraham Newman, I think, would probably find the failure of INSTEX unsurprising.  It can be very difficult to overcome the structural advantages created by a strongly hub-and-spoke network structure. 

IV.       The PRC’s “Leverage Web” Strategy

So I think the Farrell and Newman construct really can help illuminate some important aspects of modern diplomacy and international relations.  And I think it really can provide insights to inform national strategy. 

Indeed, I would argue – and have argued repeatedly in print, including in a piece last fall in the National Security Law Journal – that the deliberate cultivation of positions of asymmetric network advantage, and of the institutional capacity to exploit them, is absolutely central to the grand strategy of the People’s Republic of China.

As I see it, the Chinese Communist Party’s approach to social control is based upon a fairly sophisticated conception of how to influence populations into desired patterns of behavior by shaping their incentive structures so as to rely as much as possible upon autonomous choices, rather than upon issuing specific commands.  But to do this, one needs two types of tools: “tools of discipline with which to shape societal actors’ incentives by rewarding desirable behaviors and punishing deviant ones,” and also a system of surveillance that gives the would-be controller “a reasonable likelihood of being able to tell who is conforming and who is not, so that such rewards or punishments can be applied to them as needed.”

My recent law journal article focused more upon the “surveillance” side of this equation, but the “punishment” one is just as important because it provides the “muscle” behind the CCP effort to build an “incentives-based system of trained conformity.” 

At home in China, the massive apparatus of PRC state power and CCP influence provides those tools of discipline.  In its most brutal forms, this routinely includes arbitrary arrest, torture, physical coercion, and even such things as mass internment camps for religious minorities and the politically undesirable.  In more subtle forms, China’s evolving “social credit system” aspires to reorient essentially all of the positive and negative incentive structures of day-to-day life for ordinary Chinese people around not doing things that the ruling Party deems undesirable.

As I see it, the CCP is also interested in expanding the reach of this system of control – by degrees, little by little – into the rest of the world.  It is, for instance, working to export an embryonic version of its domestic censorship abroad, with economic and diplomatic pressures increasingly exerted against foreign individuals, companies, celebrities, law firms, or even entire countries, that dare even to say anything the CCP finds distasteful. 

Anyway, the reason I think Farrell and Newman are relevant in this context is because China seems not merely to be exploiting existing international economic dependencies upon China in service of this goal.  It is also deliberately working to create more such “leverage webs,” with strategic influence very much in mind.

The architecture of China’s signature international effort of the last decade, for instance – the “Belt and Road Initiative” (BRI) of infrastructure projects – is fundamentally bilateral and asymmetric, doing more to link BRI partners to China than to help them build any kind of independent connectivity amongst themselves.  For this reason, it also inescapably leads to ever-deepening relationships of dependency for smaller countries upon the PRC’s huge economic base.  BRI projects may, or may not, actually boost the economic growth of China’s infrastructure partners, but they always increase those partners’ dependence upon China. 

As my Hoover Institution colleague Elizabeth Economy put it in her 2022 book The World According to China,

“China has tried to portray the BRI as a multilateral initiative. Yet the reality is something quite different.  It is a collection of often opaque bilateral agreements signed under a Chinese framework notion.  The Belt and Road Forums further enhance the impression of Chinese centrality: heads of state travel to China to seek deals as supplicants to China.”

Such asymmetry, I think, is the whole point.  There may sometimes be some mutual economic gain involved, but the key – from China’s perspective – is to ensure the kind of dependencies that create webs of leverage of which China can thereafter take advantage.

As the CCP explained in its 14^th Five-Year Plan, China also aims to create what it calls a system of “dual circulation,” through which increasing demand for manufactures within China will at least partially substitute for foreign exports as a driver for PRC economic expansion, even while foreigners remain highly engaged in the Chinese market. 

In other words, both the BRI and “dual circulation” are concepts under which the CCP hopes to minimize China’s dependence upon foreign trade, while still maximizing the dependence of the rest of the world upon China.  According to the 14^th Five-Year Plan, for instance, it is one of the objectives of “dual circulation” to give China leverage over its trading partners.  Specifically, the plan says, the aim is “to accelerate the cultivation of new advantages” that are not just to be used in international cooperation, but also are to be used in “international … competition.”

Especially read in conjunction with China’s ongoing efforts to create dollar-independent trading and financial networks to inoculate it against international sanctions, this deliberate cultivation of global networks of leverage is something of which U.S. and other leaders elsewhere in the world should be more aware than I suspect they are. 

Whatever the ultimate results of China’s current efforts to manipulate the topography of global trading, financial, and infrastructure systems, I think one can clearly see a pretty sophisticated theory of network-based social control behind PRC grand strategy.  More importantly, for present purposes, I believe you can see an appreciation for the “weaponizability” of interdependence that Farrell and Newman would surely themselves recognize. 

V.        Conclusion

In conclusion, as a former senior U.S. diplomat and national security policymaker who is now trying to make his way as a scholar and educator, can I say that simply being aware of these dynamics necessarily give us a clear policy agenda and path forward as we mull over contemporary issues such as how to manage supply chain, investment, and other trading relationships with the PRC?  Alas, not. 

But I do feel confident in saying that our leaders can’t be good stewards of either our economic or our national security interests if they don’t understand dynamics and challenges related to the weaponization of economic interdependence. This is – to my eye, at least – an area where the theoretical and the practical can inform each other very usefully, to the broader benefit of sound policymaking in a troubled and often quite threatening world.

I think there is rich scope both for more research into these dynamics from a scholarly perspective, and for a much more systematic policy agenda of exploring how players maneuver for network advantage and how they work to exploit that leverage in strategic competition.

-- Christopher Ford