On Rome and America

I recently finished listening to Mike Duncan’s excellent “History of Rome” podcast series. It left me with a confusing swirl of thoughts that I wanted to put down on paper before they gradually evaporated.

One of the themes I found myself reflecting on was the Moldbug point that a government is just a corporation. Moldbug used this argument to note that the difference between how the average westerner thinks governments should optimally be structured (with democracy and separation of powers and an independent judiciary), is radically different than how they think private corporations should be structured (with dispersed shareholders who vote for a board, who appoints and supervises an all-powerful CEO). Which either means that you need to explain what exactly it is about their different tasks that justifies the different form, or  if there's no fundamental difference in tasks, at least one of the two is suboptimal. 

When you see the various governing forms that the Roman Empire took over time, it brings to mind the more basic question – if a government is just a special case of a firm, well, what exactly is a firm, anyway? And Ronald Coase (and Jensen and Meckling) has an answer for us here. A government, like a firm, is a nexus of contracts. That is, a government is a way for various people to coordinate their behavior for some overall purpose. Some of those people are coordinated against their will, which is an odd form of contract, but not a fundamental obstacle to the idea. And secondly, most contracts are incomplete. We either can’t perfectly commit to future actions, or can’t observe other people’s information sets enough to know if they’re screwing us (is their lousy performance because got a bad draw, or because they stopped putting in effort?) or a variety of other problems in contract theory.

And in this lens, to say that contracts are always partially incomplete is another way of saying that power is always somewhat informal. The org chart can specify who answers to whom over what questions, but unless you can actually enforce action the whole way down the chain of command, there’s always a “run it up the flagpole and see” aspect to any given order. 

So in this sense, the simplest version that solves this incomplete contracts problem is the early Roman Republic – when “the Romans were like brothers, in the brave days of old”, in Macauley’s phrase. With enough mutual love and regard, the incomplete contracts problem gets largely solved, through sacrifice, shared understanding, and desire to not maximally exploit everyone else. Government as the expression of self-determination for a well-defined ethnic group with mutual fellow-feeling thus makes sense as a reliable way of solving the contracting problem. 

This is like the highly successful family business. But a firm where you only hire family members is limited in its capital and ability to expand, unless you’re having bazillions of children each generation. This creates incentives supporting the tendency to hire outsiders, but keep control and cash flow rights to the family. So now you go from the family-only operation, to a family-owned firm employing outsiders. This can be quite stable and allow for a lot of growth. But a lot of firms find out that if you really want to expand to take advantage of profitable opportunities, you may need to dilute control by selling shares in capital markets, and hiring competent external managers instead of just your eldest son, or even your most competent son. 

At each step, the challenge is figuring out how to coordinate the increasingly complex nexus of incomplete contracts that holds the whole thing together. And for the case of governments, this is an order of magnitude harder than for private firms, as there’s no larger governing apparatus or courts to appeal to in order to resolve disputes. You can get some progress due to personal loyalty, or traditions being a Schelling Point for behavior, or displays of force making things collapse to a single Nash equilibrium. But if people disagree enough on what the contract requires in a given high stakes situation, maybe the nexus itself breaks, reverting to smaller groups with less ambitious aims who can still agree on things. Assuming, that is, that those groups can actually operate independently of the broken off group. Complexity, like entropy, is easier to add than it is to get rid of again.

And this sounds a lot like the early history of the Rome. In the beginning, the Romans were literally the citizens of Rome, who were also the soldiers. And their enemies were other nearby cities – the Samnites, the Latins, and the rest. But once defeated, these become subject cities, then friendly allies, and then finally, Italians, the inner core of the expanded polity. The concept of who is “us”, at the core of the high trust solution to the incomplete contracts problem, gets expanded. And you see the genuine aspect of ethnogenesis – what exactly makes a people, anyway? This concept has been so abused by blank slate progressives advocating for open borders (where "there is some flexibility here" gets substituted for "there is thus infinite flexibility here"), that it’s easy to miss that there are genuinely multiple ways to construct identity. And the surest way you can tell this is that the old group names just stop appearing in the stories at a certain point. We stop hearing about the Samnites, then we stop hearing about Hispania, then finally we stop hearing about “the Gauls” (as a people) and just hear about “Gaul” (as a province). 

The flip side of this expansion, however, is that it becomes harder and harder to articulate an operating principle of what and whom the whole thing is for, especially in terms of the question of who gets to be in charge and who gets to enjoy the benefits. The Roman consuls first have to be Patricians, but then that gets discarded, and they can then be wealthy Plebs too, but they still have to be Roman. Then once it moves to emperors, they eventually can be Roman families from the provinces, like Trajan, but still generally of Roman stock. These expansions prove successful, then you get the series of great Ilyrian emperors, like Diocletian. But then you get peasant emperors like Maximinus Thrax, and then you get Gothic generals who wield lots of power over puppet emperors, but still do good service to the empire (like Stilicho). But then this gives way to barbarian generals like Ricimer, who notionally still serve the decaying empire successfully, but historians start to argue whether he’s actually just a King. And finally Odoacer decides to make our lives easy by actually sending the purple cloak back to the Eastern Roman Emperor and saying he’s done with the whole thing, and now is just the King. Which is lucky for us, only because it makes the whole narrative simpler for “when did they meaningfully stop being Roman Emperors”. The Senate kept meeting until the 7th century, even as it was totally useless and a joke. 

And it’s not just the leaders – by the end, when the western Roman empire is fighting off increasing barbarian invasions, the legions doing the fighting are mostly barbarian too, and there’s not always a lot in the legitimacy stakes to make you prefer the notional emperor(s) from the usurper. It’s certainly not obviously the extent to which they represent “the Roman people” in a Wilsonian sense. 

If all this sounds like it becomes a confusing mess of an org chart by the end, it is. Part of this is because when we refer to this as a nexus of contracts, there really aren’t any actual formal contracts anywhere once the simple ties of ethnic kin aren’t the main driver. The closest we get is clear rules of succession, whose importance becomes very obvious. And when these get ignored, we’re left to the coordination game version of the Melian dialogue – the strong (and their nexus of contracts) do what they will, and the weak either join the winning team, or get killed. This often happens surprisingly fast, and a lot things that seem like they’re going to turn into massive civil wars end up as a total collapse by one side without a fight, or after a small initial battle, where all the supporters of one side abandon their leader to join the opposing faction, and the losing general either gets killed by his former supporters or commits suicide. Say what you will about the late Romans, but they played politics for keeps. 

If democracy has any virtues, it’s certainly not the wisdom of the people, or even the legitimacy that elections confer, but probably just the well understood rules of succession, whereby power is transferred on schedule, and the losers generally get to live out their lives in peace without being thrown in jail or killed. Then again, this was true in Republican Rome too, until one day it wasn’t. In the early parts of the story, you hear mostly about cities and structures, with the individual Romans playing less of a central role (with occasional exceptions like Scipio Africanus). By the time you’re routinely hearing about individual personalities at high frequency, you’re well on the way to the collapse of the Republic. And at first this takes the form of utilizing factions of Romans violently against other Romans, then it becomes using foreign auxiliaries to overcome Roman legions (who are themselves mixed), then it ends with the foreigners just running the show. 

In this regard, what you’re trying to coordinate with the early Roman Republic is a form of group selection, whereby members of a group all end up benefitting in evolutionary fitness terms because of cooperative action. Like ethnogenesis, progressives love this idea, and abuse the concept mightily to argue for all sorts of nonsense. But it can exist – not only in the case of a termite mound, but also in terms of the cells in your body. They engage in all sorts of group coordination so that their DNA gets transferred on. For a while, so did the Romans, and then later on, so did the Italians (at least in our metaphorical sense, though perhaps in a literal one too). But even inside your body, defection is often locally profitable without strong enforcement mechanisms. When one section of cells starts growing uncontrollably at the expense of the rest, this is called cancer, and has a lot of parallels to a rebellion in civil society. But the metaphor is imperfect, because the cancer of rebellion doesn’t always kill the host. Rather, at some point it starts to look less like the body with cancer and more like a cancer in charge of what used to be the body. The polity, unlike your body, can theoretically absorb outsiders. 

Aside from these curious aspects about the nature of government and its change and decay, the other part I found myself wondering about were the parts of the story being left out. Not as a conspiracy or anything, but just which are the noticeably important aspects that don’t seem to get much emphasis in the Duncan retelling? 

There are two big ones. The first of these is birthrates. Obviously he’s not recounting a continuous census of fertility. But in the days of the Republic, the recurring theme is always that whenever a Roman legion gets wiped out, they just raise another one and keep fighting, refusing to give in. You can see this an indomitable spirit and resolution, but it’s worth pondering the sheer logistics too – there must have been a large excess of tough military age males sitting around at any time, ready to be brought into military service. Where those males all went by the end of the empire is not well explained. The standard view is that they were unwilling to fight, and were kept from service by rich land owners who didn't want to sacrifice their workers. But it also seems quite likely that they might just not have been around in the same numbers. Either way, it’s puzzling. 

The other piece that supports the latter interpretation is that by the time you get to the Empire (where there is much more focus on individuals, so we know their life stories better), you can’t help but notice just how few children the Roman emperors seemed to have. If you’re not constrained by resources, why not have 12 children, or 20? How on earth do you reach the stage that you don’t have enough sons to continue your dynasty, when this is the single most important aspect of succession? Or only one son who's a total muppet? Weirdly, having no children at all is more understandable than only having two, as the former might just be general unlucky infertility, but the latter suggests it's possible, just not done much. I had always viewed the overwhelming knock on Marcus Aurelius as being that he broke the tradition of the “Five Good Emperors” of appointing as successor the most competent man, rather than their blood offspring. This is true, but the part that Duncan notes (which I hadn’t known) is that none of the others actually had any blood sons! Which is bizarre, when you think about it. It’s not even like they’re Henry VIII, producing unlucky daughter after daughter. Aurelius appointed Commodus, who was a disaster, but as Duncan notes, he was rather in a bind. Either he probably had to kill him, or make him successor, since letting him just hang around was likely to lead to civil war, which is much worse. Even Augustus, the most powerful man in the empire, ran out of sons. Like, just get some mistresses! Or bang your wife more! I have a feeling something very odd was going on with Roman birthrates, which I want to understand more.

The other even less remarked on piece is Roman engineering and technology. Occasionally you get snippets of the story like Trajan building the longest bridge for the next thousand years, or Caesar shocking the Germanic tribes by building a bridge to transport his whole army across the Rhine before the barbarians knew it had happened. Or the general impressiveness of aqueducts, which managed to slope ever so gradually over miles and miles to deliver water from one place to the other. 

Nobody else was doing that. So how did Roman technology advance? How late in the story did that stop happening (surely an important question to those who judge present civilisational health by the existence of iPhones). And why couldn’t anyone else match it in the same way? You realize that you have such a tiny number of primary sources, and if it wasn’t a subject that people bothered to record the history of, you’re going to have a lot of guesswork.  

Finally, comes the applications to the present. One aspect that stands out is that the ability to forecast the lifespan of the empire is complicated by the fact that battles at that time were still largely tactical and not strategic. That is, there often might be only one or a couple of large encounters that decide who comes off the best in a war between nations. And because these often turn on small decisions on the battlefield, smaller armies can frequently rout and destroy much larger ones, or a small number of surprise defeats can threaten the entire nation (like in the Punic Wars with Carthage). This means that at a lot of places, something going differently could have ended the whole Roman experiment much sooner – the Punic Wars, for instance, or the Crisis of the Third Century. In the age of strategic warfare, you might actually have a better chance at forecasting which structural forces will prove decisive if the two sides just grind it out long enough. Then again, maybe not – there’s been no shortage of American defeats to notionally weaker foes that might not have seemed structurally obvious ahead of time. 

The last one is the surprisingly low correlation between the level of domestic misery at any given time and the actual instability of the whole project. The late Roman Republic was an utterly miserable time for its citizens compared with the previous centuries, with the advent of massive Roman civil wars, proscriptions and deaths of political opponents and their supporters, and general chaos and uncertainty. But all the protagonists at this point are still Roman. You never really get a sense that there’s an Alexander of Macedon equivalent waiting in the wings to crush all the war-weary parties. Instead, Rome still seemed pretty secure from external enemies. By contrast, there are lots of periods during the Empire where the barbarians seemed to be both stronger and more coordinated, and could tip the whole thing over. Even here, it’s not an inexorable decline. By the early 5th century, things actually seem to be looking slightly brighter for a while. Yet the striking thing at this stage is that even a competent Emperor like Aetius isn’t able to actually fully reverse the losses, only slow them. It’s not long then before things go off the rails for good, at least in the west. 

If we look at the level of America’s foreign predation, we see useless and counterproductive wars, and endless third world migration, but there is nothing like barbarian armies forcibly peeling off territory, or menacing the homeland. This calculus is made a little more difficult by the fact that we are highly dishonest about what the American empire actually is, and thus what it would mean to be losing territory if you only rule indirectly. If you defined the American empire crudely as “everywhere that had a George Floyd protest in 2020” (including places like Berlin, London and Melbourne) then maybe the calculus gets a little harder. But still – in physical fights, America loses, but it loses in other people’s backyards, when the stakes for the domestic citizens are mostly pretty low. 

And in a pinch, I find myself thinking that our recent developments seem to map more closely to the late Roman Republic than the late Roman Empire. If I had to guess who Trump most closely matches too, it’s probably the Gracchi brothers. They started using populism, and in their case threats of mob violence, to get their legislation passed. This isn’t exactly Trump, but the sense of tapping into populist anger to circumvent the normal political process seemed like a good description of his nomination and victory in 2016. And the establishment backlash to unseat him, both from dubious Russiagate investigations from the civil service and Democratic party, periodic deaths and violence in political confrontations and protests becoming more normal, unusual levels of election fraud in 2020, and Trump’s final temper tantrum protest on January 6th, all fit the same pattern. A gradual erosion of previous norms, a gradual escalation by each side. The Gracchi brothers both got killed, which I don’t think will be Trump’s fate (though blog.jim still is betting on this). But the important parallel I see is that they showed a path to a certain kind of escalation that later men like Marius, then Sulla, then finally Caesar and Augustus, would both increase and take to its logical conclusion. At some point, people just realize that the Republic is dead and not coming back. It takes a long time, and a lot of denial, before that point is reached.

Sometimes, that’s just your lot, and there’s not much you can do. To modify Brad Pitt’s character in Inglourious Basterds – you don’t gotta be Niccolo Machiavelli to know that you don’t want to live through “the Year of the Four Emperors”.

But America as the late Roman Republic is actually the optimistic scenario. Not because it won’t get worse – if the metaphor holds, it definitely will. But because it might eventually get better.