Center for Strategic Communication

For the second in my interview series, I interview newly minted political scientist W.K. Winecoff. W.K.’s work is in network analysis, an area that many security thinkers are increasingly embracing. But he also does International Political Economy (IPE). Cash Rules Everything Around Me (C.R.E.A.M) as DoD is currently discovering, so IPE work is an area many security thinkers not usually inclined towards dollars and cents should check out. Lastly, he’s also one of the O.G.s of the political science and International Relations blogging team. I thought I’d ask W.K. a few questions about his work. It turned out that W.K. is quite talkative when it comes to this stuff, and I’m posting his interview excerpted. You can read the full, uncut interview here

Follow W.K. on Twitter or check his homepage.

Adam Elkus: Lots of people in national
security and international relations are attracted to network analysis. What
prompted you to study networks? 

W.K. Winecoff: With the proviso that my work (much
less my philosophy) is not all-networks-all-the-time, I think the best answer
is that network analysis conforms to many of our intuitions about the way that
social systems work. It is relational and systemic, interdependent and
hierarchical. A society is a network; that’s why we often talk in terms
of community, polity, market, congregation, corporation, whatever. These are
ways of describing linkages between people and groups, which is the definition
of a network. So why not use network tools to study these systems?…..

{I]n IR and IPE we’ve
spent a lot of time constructing elaborate theories related to complex
interdependence, spatial dependence, diffusion/contagion, joint memberships in
institutions, epistemic communities, and systemic relationships more broadly.
Network methodologies are very well-suited for working in those contexts.
Network science has now advanced to the point that there are some pretty
well-established methodological tools — inferential and descriptive — that
folks who understand basic probability can more or less pull off the shelf
without too much trouble. Until very recently that hadn’t been true, so I
understand why network analysis is still catching on in some circles, but at
this point there aren’t any good excuses for not using these tools when its
appropriate to do so. And it is almost always appropriate to do so……

Perhaps more importantly, other methods are often not
well-suited for the analysis of world political systems. Regression-based
statistical models assume that observations are independent from each other and
are distributed identically (I.I.D.). If this is assumption is not true then
statistical results will be biased in an unknown direction by definition. I
know of no substantive theory in IR, IPE, or really any other corner of social
science or foreign policy analysis that would accept that assumption as being
valid. Not a single one. Sometimes that might not matter: we’ll get lucky, the
dependencies will be weak, the errors will still be random, and we won’t commit
an inferential error. But it’s heroic to believe that will always, or even
usually, be the case……

Closer to home for me, the financial
crisis revealed the importance of this in two ways which have clear analogues
with many other areas of IR, security studies, comparative politics, and other
fields. First, because no quantitative IPE folks saw the financial crisis
coming. We had no models, no empirics, basically nothing. We didn’t even have a
way to contextualize it. It was like the end of the Cold War for IR, 9/11 for
security studies, and the Arab Spring for comparativists. We all missed these
huge events, which were clearly a spasm of interdependent processes.

Second, because no quantitative IPE
folks were able to say anything much about the crisis after the fact — and
security folks and comparativists spent some time flailing after 9/11 and the
Arab Spring, respectively — except what everyone else was saying: bankers are
greedy bastards; blowback (i.e. “balancing”); autocracy isn’t
sustainable. So what? Bankers are always greedy but we don’t always get global
financial crises. The U.S. has had a foreign military presence for 70 years but
we never had a 9/11. Autocracy has been the norm in the Arab world for decades
but we’ve never had a cascade of revolutions…..

Now we’re starting to go back and
revisit some of [standard IR/comparative politics] claims using network methods and we’re finding out that
many of them aren’t robust at all. Skyler is leading the way on this in
security studies. He’s already called the democratic peace into serious question.
Not any particular theoretical explanation of it… the correlation itself. Once
structural network processes are taken into account a lot of the variables at
lower levels of analysis wash out, including regime type. The same is true in
other areas. This is the danger of the I.I.D. assumption: it can’t handle
endogeneity, and a lot of what we observe is endogenous. I have a pretty strong
suspicion that analyses seeking to explain phenomena by reference to regime
type are pretty much over. If I’m right, this will be the biggest development
in the discipline in a long time……

AE: What do people most frequently misunderstand about

WKW: In terms of IPE — and to a lesser
extent IR — I think there are two. First is that network analysis is just
about drawing pretty pictures. That it’s not rigorous empirically. Whenever I
hear that — and I still hear it a fair amount — alarm bells start ringing.
It’d be like saying that cosmology is just looking at sparkly things in the
sky. That may have been true a long time ago ago, but it’s certainly not true
now. It shows a fundamental lack of understanding of what is possible. We now
have a lot of descriptive and inferential tools that make detailed empirical
examinations of networked systems feasible.

The second is a two-parter,
concerning a belief that it is not rigorous theoretically: that insights from
network science are trivial and/or obvious. These are related but distinct. I’m
not quite sure why people think networks are trivial in IR/IPE when no one else
in the world seems to, but they do. I think it’s because they don’t understand
the importance of topology and dynamics in a network context. At a prominent
panel at the flagship conference of the International Studies Association, I
heard one of the most famous internationalist political scientists say, with a
straight face, that we all know that networks are important because if one node
gets infected any other node can get infected. That the only insight from
networks was like something we could learn from the movie Contagion.
I’ve seen that movie. S/he described it well. But that is fundamentally not
what network science says. Not all network topologies behave the same in the
face of shocks. Not all pathogens (or financial crises, revolution diffusions,
etc.) spread the same way. So the structure is not trivial, and those who think
it is do so because they don’t understand quite what they’re dealing with. (As
an illustration, this scholar was using the Contagion example to argue against
the importance of structural theories: yadda yadda things spread we all know
that who cares

Another example: one of the most
cited papers in IR that mentions networks distinguishes them from
“markets” and “hierarchies” as distinct organizational
forms. This, again, misses the point entirely. Markets are networks:
they are an aggregated grouping of interdependent actors which are connected by
the relationships between them. And networks can be — and almost always are in
real-world social systems — hierarchical, and a hierarchical organization
cannot be anything other than a network. So I think there is some foundational
misconceptualizing going on that I’d like to see corrected.

In particular, I think the discipline
doesn’t understand the difference between simple random networks — the kind in
Contagion where all nodes and all links are basically the same — and
complex dynamic networks — the sort that constitutes the global economy, where
all nodes and links are demonstrably not the same. The two types don’t
behave the same way. They don’t have the same properties. They can’t be treated
as if they were identical, either in theory or method. They are so different
that in some ways I wish they weren’t both called “networks”…….

AE: You are one of the most vocal
international political economy proponents in the blogosphere. What does IPE
give a policy analyst or political scientist that other approaches lack? 

WKW: Thomas Oatley (my dissertation
advisor) has argued that at best academic social scientists are irrelevant to
policymakers. His basic point is that we generally seek to explain why they do
what they do. Presumably they already know why they do what they do, so why
would they need us? Perhaps this is why economists have a more prominent role
than political scientists. I disagree with Oatley, to some extent. I think the
information problems policymakers face are significant enough that IPE should
be able to help provide policy advice.

But I don’t see that as the central
task of IPE scholars. Instead, I see our role as more critical and our audience
as the broader public. Economists tend to assume that Benevolent Social
Planners are trying to set policy optimally; when done well, IPE demonstrates
emphatically that that is not the case, which is why the world does not work at
all as economists believe it should. We tend to emphasize the importance of
power when setting policy, highlight rent-seeking, and generally take a
critical line. We can also illustrate structural features of the world economy,
and explain how the structure confers power, whereas economists focus almost
entirely on agents which they perceive to be autonomous. So I’d say that IPE’s
comparative advantage is showing how the world actually works while
economists argue about how it should work.

I think this leads (or should lead)
IPE away from the moralizing that economists and much of the lay public quickly
fall into. If interests and ideas compete to set policy then there is little to
moralize about: one side wins and the other loses, but both are trying to
advance their interests. Some IPE folks side very clearly with some social
class or other group, but this is an aesthetic choice. There’s no moral reason
to do so, or at least none that comes from our discipline. In this way IPE is
much more like “science” than economics, because it is more grounded
in empirical reality than just-so theory. Economics believes in a social
welfare function; political science does not. Daron Acemoglu, an economist who
steps more and more into IPE and CPE (comparative political economy) as time
goes by, recently co-wrote a very interesting working paper excoriating the
economics discipline for just this thing. He’s basically saying that economists
need to become political economists, which is what we in IPE are already
supposed to be doing. I went through the same transition as an undergraduate
economics major, which is why I went to graduate school for IPE rather than

I quite like the amoral character of
IPE. Paul Krugman constantly says that “economics is not a morality
play” but he’s one of the biggest moralizers of all. In fact it’s hard to
find an economist that isn’t ideologically committed. In my opinion this tendency
has a very negative impact on critical thinking. For example, Krugman wrote a
big essay in the New York Review of Books accusing Ben Bernanke of being
brainwashed by the Fed “Borg”. For Krugman this was the only possible
explanation for why Bernanke wasn’t following the Optimal Social Policy. I
couldn’t believe that such a sophisticated person would make such an
unsophisticated claim in a prominent media outlet. An IPE person would never
write that way. We’d look at the ways in which interests and ideas are filtered
through institutions to explain why policy is set the way it is. We don’t think
there is any such a thing as an optimal policy. That’s pure fiction. Just like
the Borg……

AE: You’ve also had some beefs with
poli-sci bloggers over the years over war, peace, and international and
domestic economics. What do you think your most contentious blog post

WKW: All my posts are contentious. My
attitude towards the blogosphere — and, to some extent, academia — is that it
serves no purpose if it’s not about contesting ideas. And when ideas are being
contested the stronger the better. Come direct or go home. Be prepared to be
wrong, too! There’s no shame in that. I try to learn something every day, which
means that every day I realize I was wrong or ignorant about something. It’s
not a big deal, it just requires being realistic. I don’t know why sheepish
people get in this business, and I don’t understand why people won’t update
their beliefs when they do learn. The dialectic remains important.

Anyway, I write in a provocative
style quite often so it’s hard to know what the most contentious post was.
Sometimes people don’t notice or just let it go, but in terms of attention received
the most controversial had to be the one where I said that Paul Krugman
has the political sophistication of a 6 year old. Which is true, as an
exaggerated description. He’s all about blaming others; he’s faultless himself.
He’s a truth-teller and everyone else are liars. He’s been sticking his tongue
out at everyone for years on the authority that his preferred policies would
have worked had they been tried, but they weren’t, things aren’t great, so…
there? He’s been running victory laps despite having been defeated in every
campaign he’s run. This is childish, I’m sorry.

 He could be 100% right about the
economics but he should really just stop writing about politics. He’s in way
over his head. It’s a somewhat-regular theme for me, or used to be. I’ve laid
off recently. But one of these times Henry Farrell took exception at Crooked
, and it got picked up all over the place: Krugman’s NY Times blog,
Drezner at FP, Nexon at Duck of Minerva, Daily Kos, everywhere. The CT comment
thread was pretty epic even for CT. The consensus seemed to be that I was
either a fool or a knave because my argument did not reduce to “evil
Republicans ruined the world on purpose” (see above about how the
amorality of the IPE approach appeals to me). I believe only Phil Arena and Dan
Drezner were on my side.

But it never bothered me. I still
think I’m right on the merits, and moreover I was arguing from the most
left-wing position of anyone. Nobody got that, and so I was being criticized
from the left for my criticism of the bourgeoisie(!). My argument was Marxian,
it was structural. Many of my arguments are, even though I’m not in any strict
sense a Marxist. I’m sure I could have come across better but I don’t think it
would have mattered. In some circles if you attack Krugman it’s knives out.
(That’s another problem with the doctrinaire left these days: if you can’t
criticize the op-ed page of the New York Times, then who the hell can you

Anyway, that would have to be the
biggest controversy. But I enjoy arguing with people smarter than me as a
general rule. I’ve been fortunate enough to have attracted the attention of
some of the blogosphere’s leading lights, and they’ve been kind enough (or
defensive enough) to engage with me. I’ve seldom come away from those
discussions without having improved my understanding, which is my whole goal. I
hope I continue to have opportunities to participate in those kinds of