Probabilistic Minarchism

A survey of Adam Gurri’s conception of Probabilistic Libertarianism and a permissionless society.


Taleb’s Probabilistic Minarchism

Writer Adam Gurri imagines Taleb’s political philosophy, taking into account his attitudes and arguments from Antifragile. First, he makes a good point here, and one I’d probably prefer to regurgitate when asked about my own ideas of an ideal government:

There are inherent problems when talking about ideal governance arrangements, of course, which may be why Taleb has mostly avoided it—reality constrains you both by how hard it is to get to where you want to be from where you are, and by what you would set in motion if you were even able to get to your ideal.

I might also aside here and say that Gurri’s ability to make the claims he makes in his articles don’t just speak to his own insight but also to the brilliance of Antifragile, which only really enjoyed a lukewarm response from a lot of reviewers that I respect. I thought Antifragile was masterfully written, such that I got some sense of a characterized fiction-Taleb, one whose response to the news I could probably predict, since I had some sense of his preferences and aversions. I know how this character likes to eat and exercise and what he would likely think about issues that weren’t touched on in the book. I think this is what Taleb was attempting first in smaller doses with his earlier characters like Fat Tony. I thought his asides made some sort of sense in inundating the reader with this idea so that it wasn’t just restricted to “this economic idea” in the Econ container of the mind.

Anyway, Gurri argues that Taleb would likely be a Minarchist, but for “probabilistic” reasons instead of any of the classical Minarchist arguments.

The argument of the book Antifragile is that when you have a lower bound on bad outcomes, volatility can only be good for you, as the most extreme variation can only occur in the direction of good outcomes. As such, it should not come as a surprise that Taleb believes in a strong government role for bounding us against tail risks such as being conquered, having a rogue GMO wipe out all of agriculture, or permanently damaging the environment. Other than that, he believes that our policies should be as libertarian as possible. Bound the negative outcomes, unbound the positive ones; often the opposite of what governments actually do today.

At first glance this appeared even more extreme than your typical minarchism, as minarchists also believed in a role for the state in protecting property rights and policing violent crimes. Here, Taleb is not anarchist but rather something like an extreme federalist; such things should be left to local governments. Indeed, if Antifragile is any indication, it should be even more local than we are presently willing to consider; we should be more like the Swiss, atomized into a large number of tiny cantons run on the basis of direct democracy. This is another way of bounding negative outcomes—any extremely bad local policy can only have an impact on a very limited number of people.

[more here]

I looked around for more on Probablistic Minarchism, but (by that name, anyway), no one really expounds on the concept. So I returned to the author, Adam Gurri. Sure enough, that was not the first time he mentions probablistic libertarianism or connected Taleb to the philosophy.


Towards a Probabilistic Libertarianism

Adam Gurri recently posted the question “What is government?” in a recent his Umlaut column.

There is no true central decisionmaking agent known as “government”. There is an ecosystem of institutions that are funded entirely with tax revenue, which is collected under the threat of force. The diversity of these institutions cannot be understated—common law courts have a history dating back centuries, perhaps millenia. The Department of Homeland Security, on the other hand, can claim just over a decade, and it ropes together a group of institutions supposedly under its direction that predate it by decades.


As with “government”, calling this hodge-podge “the market” or “civil society” is not very informative. These categories are so broad as to rarely be very meaningful; we would do better to delve into the strengths and weaknesses of the specific institutions we have in mind.

Which is a point I’m surprised I haven’t really stumbled through thus far. In his August article “Towards a Probabilistic Libertarianism” he acknowledges the libertarians’ problem in “ignoring local bullies” but insists that they are better than allowing a “single big bully” for reasons of risk. Coupling more and more factors into a single fail point allows for massive scale failures that local bullies could never afford.

Human social systems are ever-changing and probabilistic. New innovations—in a broad sense of the word which includes norms, technologies, and fashions—are constantly being filtered out or diffusing at a variety of scales within a given social system. Big discontinuities do happen, and are ultimately what drive truly historic changes, such as the spread of modern democratic institutions or the explosion of wealth in the past 200+ years. But bulk of the changes that happen within a society are path dependent, determined largely by how centuries of previous diffusions shaped present institutions.

We cannot predict this process of diffusion; a ton of money and brainpower has been spent in a failed attempt to do so for marketing purposes, never mind predicting the larger movements of society. However, we can think probabilistically about comparative institutional arrangements, and what the possibilities and limitations are within each of them.


It’s this probabilistic reasoning that has made me weary of libertarians’—or anyone’s–attempt to simply get better laws passed in order to improve our circumstances. Long term, lasting change requires shifting the probabilistic structure of governance. This happened once when the Industrial Revolution gave us what Eli calls the technologies of control—big producers who employed huge swaths of the population in densely populated metropolitan areas. This and the resulting wealth made the rise of the modern central state possible, and highly probable to emerge in some variation across the world.

It is for that reason that I also think that libertarians are more likely to affect change byattempting to engineer it, though I’m skeptical of the odds on that front as well. For the most part, I think we’re unlikely to arrive at a better equilibrium except by the same means that we arrived at this one: by chance.


The Ratchet Effect

Another interesting phrase in the this libertarian’s toolkit, providing a “probabilistic theory of the growth of government”.

Institutions like the NSA gained their clout, their budget, and their ability to attract top talent through entrepreneurship; though the creatives energies of this activity are directed rather differently than their private counterparts would be. Robert Higgs described the ratchet effect, whereby the individuals at the head of such organizations sit and wait for opportunities—such as big media events presented as crises—to expand their scope. But you need not buy into any assumptions about the kinds of people that typically run these organizations. Even if variation in governance across history and countries is just a big game of roulette, it’s a game rigged by the ratchet effect.

Start with a simple model of institutional variation. Some individuals in public office have their authority and their funding expanded, and some do not. As with private success, it is impossible to tell how much of this success is due to skill and gumption, and how much is due to simple chance. Some of these expansions of authority involved creating new titles or institutions that persisted beyond the life of the individual. Over time, this story repeats; the dice are rolled again and again. Some titles grow into institutions, some institutions gain increasing autonomy from the rest of the government. Those institutions that do not expand, stagnate, those that do, benefit from preferential attachment—and are more likely to expand yet further.


From this point of view, the question is not why government has grown so big, but why it did not do so in general prior to the 19th century. Eli favors a technological explanation; and certainly it seems that the tools of mass production and mass media were more vulnerable to central control, and also increased the ability of central governments to exercise their authority across great distances. An addendum to this that I would add is that we simply grew more wealthy—governments are constrained by real resources just as private actors are. More resources in the economy meant more resources that could be captured—and thus the process of variation that generates public institutions could scale up just as the rest of material human life was scaling up as well.

Note: I find it plausible that the answer to the “why didn’t this happen sooner” question is slavery and the political dynamic it engendered.


Two Visions of Open Societies: Voice and Exit

In this article, Gurri discusses Popper’s defense of an “Open Society”, which relies on Hirschman’s Voice and Loyalty (“rational conversation and the institutions of democracy”) to dissent and mold society. Popper is “perfectly content to engage in social engineering, so long as it was “piecemeal”, and therefore allowed for clear experimentation with clear grounds for falsification.”

Popper understood that adopting rationalism was not itself a rationally-founded choice, but a moral one. He justified this adoption on the grounds that rationalism offered the only path to non-arbitrary decision making. In Popper’s world, it’s either rational debate or chaos, reason-driven decisions or knee-jerk emotional appeals. The reality, as we now know, is that it’s always much closer to the latter. To the extent that there is such a thing as “reason”, it operates very narrowly within the context provided by the people around us and the culture and traditions we are embedded within.

Piecemeal social engineering, as Popper envisioned it, has never existed in this world. In a democracy, there are far too many interests involved in any such project, poisoning the well of knowledge before the first drop is drawn. Even in an absolute dictatorship or totalitarian government, holding power is only possible by keeping together shifting coalitions which have interests of their own. Scientific, piecemeal social engineering always devolves into regular old interest-driven, unscientific social engineering.

Gurri offers an alternative vision of an open society to Popper’s Voice/Loyalty Open Society, which is a permissionless-default Exit-option society.

The “open” in this vision comes from tolerance of divergent practices. Its virtues include increasing the diversification of our practical knowledge, and the simple dignity of treating a citizen like an adult who is capable of making their own decisions. Having a group such as the Amish who live wildly different lives from the mainstream of society should be seen, in light of this ideal, as a point of pride rather than a quirk of a free society. So long as the right of of exit is secured for those now embedded in the Amish communities who may wish to leave in the future.