The word and concept “empire” isn’t terribly useful. In some sense it implies a state made up of multiple smaller states, the leader being a king of kings or the like. In common use it implies a state that rules over people who don’t want to be a part of it anymore, or is otherwise bad in our eyes; if we think the state is good we say it is a great nation or something. Neither is very useful, because all states are made up of multiple smaller power centers, and all states have groups that would like to leave and have their own states. The Kingdom of Rome had its families and tribes that were their own dominions as well as those who would like to secede and no longer be subject to the rule of another, just at a smaller scale than the later Empire.
Yet there is a useful comparison between the degrees of these issues. Empires are difficult to hold together, often falling apart after the founder dies, while kingdoms or other smaller political units seem to be the default setting for humanity. Worryingly for moderns, often a stable kingdom grows normally enough until it hits some tipping point and, whoops, everything crashes down within a few years. Success grows the state, but at some point that growth cannot be sustained, and it isn’t obvious that the state is capable of recognizing this and pulling back from the brink.
So let’s break this down and see if we can get a sense of what the relevant tensions are here. I think some of the insights will prove useful in understanding why the USA is straining so hard lately, as well as why Russia might not benefit from taking Ukraine as much as Putin might hope, or any manner of other situations modern or historical.
First, I want to define a state (empire or otherwise) as a group of people in a specific area all ruled by a specific subset of people. So the British Empire might have been spread out all over the place with many sub governments for different regions, but it was ruled from London overall, so it was a state. The United Nations has many sub governments, but it doesn’t have any unified government that can command obedience at the lower levels, so it isn’t really a state. A ruling government must be able to make demands on behavior and have its subjects follow those demands, willingly or not. If it can’t make that happen it does not really rule.
So a state is essentially defined by the highest level of political or governmental authority (the ruler) that can enforce its will on the whole polity. This makes some of the more overlapping historical situations a bit awkward to describe, e.g. the Catholic Church and various monarchies during the middle ages where both sort of could command obedience of the same groups of people, but in general it works1. What matters when it comes to the governments of states is whether or not their subjects will stay in the state or (violently) break away, which is to say whether they will obey the rulers’ demands or not.
Making demands on people’s behavior is a tricky business. Although famously “humans do not desire freedom, but rather fair masters2” the word ‘fair’ is doing a good bit of work there. People like to have leaders and bosses, but get pretty shirty when made to do things they don’t want to do, or when they perceive injustice from those in charge. So while a ruler must be able to make people do what he wants them to do to be considered a ruler, if the people don’t like what he wants they are going to resent his power and eventually try and remove it. If I were trying to publish this essay in a journal, here is where I would start the mathematical modeling, with something along the lines of “Social cohesion is a function of the ruler’s demands.” An easier way would be to say that the more crazy or stringent a ruler’s demands, the less happy and willing to be subjects of the ruler the people will be. (I will use “happiness” to mean “stability based on subjects’ willingness to follow the ruler’s dictates” from here on out, if only because years of Civilization and Total War games have driven that concept handle into my head.)
That can’t be the whole story, however, because historically we see states with very stringent rules whose subjects seem happy with things, as well as states with very lax rules that can’t seem to maintain themselves despite that. I think the key insight that solves this problem is that what constitutes ‘fair’ rule is highly subjective, and so the more diverse your population the more difficult it is to govern them under one set of rules.
Now I don’t mean diverse in the boring modern US sense of different melanin levels or reproductive plumbing, but in the much more important sense of culture, sense of justice, what defines a good life, etc. People will live with very stringent laws if those laws match what they think is right and proper, and in fact they will demand those laws, and resent laws that do not match what they think is right and proper, regardless of their skin color. If everyone you hope to rule has very similar preferences for what should and shouldn’t be allowed by law3, you have a low diversity population and so long as you only demand things that follow their preferences you can have all sorts of strictures and remain secure as ruler. If everyone you hope to rule has very dissimilar preferences, however, you can at best rule with a very light touch and remain in power, and even that might not do it if some factions are calling for more strict rules4 and you can’t push the decisions on those rules to a lower level of government.
So looking at our model again, we can say that within a state Happiness depends on Strictness of rules and Diversity of the state. If we simplify a bit for now and assume that pushing the demands for stricter laws down to lower levels is possible, we can make a simple H = -S*D equation, where if either Strictness increases or Diversity increases Happiness declines, and the state is harder to hold together. If Diversity is low because everyone wants the same rules (including the ruling class themselves) you as ruler can do whatever you want and still have happy subjects because whatever you want turns out to be pretty much whatever they wanted. If Diversity is high and everyone wants different things, you are going to have a really hard time making any rules without pissing a bunch of people off, so your best bet to keep things together is to demand as little as possible and let lower levels of government do the bulk of the work.
Whence Diversity?
Now it is worth asking where Diversity comes from. At a fairly high level I propose two main drivers of Diversity within a population: population size and the population area. I think it fair to say that all other aspects stem from these two.
Population Size
There is a lot of deviation across people for various traits, but much of the distribution is clustered around a mean. Fancy talk for “Most people are pretty much alike, but there are some real weirdos out there.” In a tribe or village of a few hundred you might only have one or two weirdos who deviate from the local norms; in a large modern city of millions you probably will have thousands of weirdos, so many that they can form subcultures within which they aren’t weird anymore. Simply by adding more people you are going to add to the Diversity of a group, in a non-linear way.
The non-linearity means that there is a qualitative difference between ruling over village of 500 and a city of 500,000, and not just being 1,000 more difficult. Whereas the ruler of a low population state can exile any weirdos whose preferences prove too annoying, the ruler of a high population state might have to exile thousands of people to achieve the same end. Although it might represent the same percentile of the population being exiled there is a big difference between kicking out a few weirdos and kicking out a few thousand representing an entire subculture, or an army division. As the population grows so too the number of subcultures with many members that can support each other against the demands of the state, and so must be appeased instead of merely forced to follow the will of the rulers.
This effect can be seen in the unofficial social norms of different sized communities as well as the official rules. Small towns have a reputation for conformity while cities’ anonymous masses contain all manner of social deviance, good or bad5. At the very least, one has to be very abnormal indeed to draw attention walking down the streets of New York City compared to, say, Corfu. Note too, that while people might not laud the wonders of the one horse town compared to cosmopolitan NYC, the amount of social unrest also tends towards zero compared to the big city. Small town folk generally know how to behave, in large part because there is a single “way to behave” that has become the norm, as opposed to the large numbers of competing ways to behave adopted by various subgroups within a large city.
So, just having a lot of people under your rule is going to limit how much you can make people do what you want before things get heated, and that’s independent of…
Population Area
Population area is, well, the size of the area your population lives in. This is a little different than the size of your country. You might have a huge state laid out like Alaska with pretty much everyone in the bottom 5% of the area and the upper 95% mostly bears, and so unless the bears start getting ideas above their station you are really just ruling the lower 5% or so. Likewise, you might face something like the British or Athenian empires where you have a lot of population centers spread across lots of ocean, but the ocean doesn’t really count because people don’t live there. At the other extreme, a smallish country in terms of square miles might have many different population centers spread about necessitating multiple forces to keep things under wraps if the populace (or bears) start getting uppity. So the official size of the country is less important than the distribution of the populace within that area; the important part is where people live and how they are distributed over the area.
Why is the size of the area the population inhabits so important? A few reasons. As mentioned above, if you have lots of different concentrations of people you need to have lots of difference concentrations of people enforcing your policies, some sort of, I don’t know, policy men. Maybe call them ‘police’ for that French flair, if you can figure out how to do the little squiggle above the e. Everyone being within a few miles of each other is really handy, as at least if the police go to the wrong house they don’t have to walk too far to get to the right one.
More importantly, however, is that the cultural needs of different living areas is going to effect the norms and preferences of the people living there, and thus create diversity in demands for policy.
Imagine you are trying to rule a country much like Pennsylvania. You have two ports on either side of the box like Erie and Philadelphia, some river cities that allow for interior transportation, and a lot of mountains and woodlands in between. Plus side, all that water makes for relatively easy transportation6. Down side, the people who live in a large sea port like Philly are very different from those who live in the ridge and valley region towards the center of the state.
Why? Well, the behaviors of people who live in cities tend to be very different than the behaviors of those living in small towns and rural villages and farmlands. As mentioned above, cities tend to be more anonymous and people interact in a more arms length nature, but also tend to be more commercially focused and interested in high culture. Rural folk tend to be a bit more practical and less interested in attending the opera say, while paying more attention to the local community members. Trade is as likely to be in kind or barter based than for cash. (My neighbor might help me repair my lawnmower for a drink and the implied promise of future help; that isn’t going to fly with a hired handyman, and offering my neighbor cash would be insulting.) As a result the nature of the norms of behavior and laws governing trade are likely to be quite different, and indeed even the divide between what is a social norm and what is a legally enforced rule will shift quite a bit.
Further complicating the matter is the macro cultural differences (for lack of the better term) that grow between populations that do not interact a great deal. The American south east is a very different place than the north east still, and the upper mid west is as well. I lived in Minnesota for a few years, and the whole “Minnesota nice” thing is no joke, outside the cities at least. Similarly, the norms around talking with strangers in the south are slightly uncomfortable to a northerner; we don’t expect people to actually want to hear our life story while waiting in line to check out7. Add in details like “Is it ok to canoe down part of a river that is entirely contained within one person’s land” or “How do we worship God in the local church?” and things can get a bit crazy if you decide to pass laws thinking “Oh, everyone is pretty much like me and the people I live near.” Hell, I didn’t even realize getting the first day of buck and doe seasons off school and work wasn’t normal till I went to college. Turns out it was only a central PA thing that everyone went hunting two days a year.
Then you get other issues that drift into the technical. We can easily imagine people living in arid or desert climates having different cultures than people living in water rich regions. What we might not consider is that even if they share the same culture they might have vastly different legal needs. A key point of stress in this case is that of water use rights. If you live in the eastern US you probably never worried much about water because it was pretty plentiful, usually just a few feet under the ground. You dig a well or tap into a river and you are good to go, use as much as you want. Really, until the stuff starts seeping into your basement or coming through your roof, no one thinks about it.
When people start moving into the western US, however, this legal framework starts to cause problems. Use as much as you want pretty quickly turns into “Oh hell, there isn’t any water left, and wasn’t there a river there yesterday?” It becomes very important who owns what water and how much volume, and the eastern model of cheap water in endless amounts just won’t work. A more nuanced system is needed.
Yet at the same time, the more nuanced system would be equally galling if applied back east. The costs of time and energy enforcing the rules that make perfect sense when water is scarce become obnoxious and burdensome when applied to nearly limitless supplies. A one size fits all solution is out of the question. You might be lucky to get away with 5 or 6 solutions that get everyone sorted out effectively8.
Making matters even worse is the question of service provision. Without getting into too many gory details, how you spend your tax revenue is as important, possibly more so, than how you collect it with populations geographically distinct enough to notice. If you tax everyone equally but spend an unusually high percentage of the revenue on your big port city to encourage trade and commerce, are the inland folks going to be ok with that? If you instead spend it on developing the rural hinterland are your urban elites going to feel short changed? Are different regions going to bicker and argue over who gets the military bases and related spending? Are your city folk going to grumble about soldiers from the uncouth mountain tribes being stationed in their town? Why do only the urban centers and suburbs get city water and sewage service when everyone in the region pays the same tax rate?
So, the more spread out in area your population is, the more diversity it creates. Ten million people living in one city with a few suburbs is very different than ten million people spread out in two or three cities and tons of little towns when it comes to ruling. If you ever wondered why in ancient times city states were so common, there you go. Likewise, there is a reason that most European states are so small. Indeed very few states are continent spanning, even today, despite low population.
Where were we…
Oh, yes, our model. Population size and the size of the area the population inhabits and its densities definitely affect the stability or happiness of the state. In our mathematical model we get something like:
Happiness = -Strictness * (PopulationSize*Area)
Maybe Diversity is closer to Population Size plus Area or something, but let’s just roll with multiplication for now, also ignoring coefficients and possible exponents while we are at it. If we really wanted to get crazy with the math we could come up with a calculation for Area based on distance of population centers and density.
I don’t want to do that. Moving on.
One ignored variable does deserve looking at…
Technology
Technology both makes the world smaller in terms of travel, communication and conformity of culture, but also allows for better surveillance and enforcement. That all seems relevant to how easy it is to get your subjects to do what you want. In modern times we have seen a unification of culture across regions in the USA due to everyone watching the same movies, the same cable news, the same television shows, the same internet sites (sort of) that would be unimaginable just 50 years ago. Regional accents have largely disappeared, to give an audible example. The number of newspapers has plummeted as local papers disappear and are replaced by national sources. You can fly across the country in an afternoon relatively cheaply, and eat at the same chain restaurant on either end of the flight, should the fancy take you. All while your location, spending and many other activities are tracked and stored for future reference.
Of course, at the same time technology is making things easier for ever more subcultures to develop, linking up people with similar interests all over the world. No matter what you are into chances are pretty good you can find some like minded people near you. Angry about what the government is doing? It has never been easier to find other people who would love to go protest with you about it. Vaguely unhappy about the state of your life? So are millions of others, and they would like to introduce you to their local cult leader/community organizer to tell you whom to blame and what you can do about it.
Is technology a stabilizing or destabilizing force, then? I don’t know. It seems that for most of the span of history technology has always pushed in the direction of larger states containing more people. The last two hundred years saw many large states emerging and consolidating, and while some have rather spectacularly split apart in that time, it seems that the general trend has been to have more people governed by fewer states. Part of the trouble with that trend is that many of the large states are a result of larger states, the British and Ottoman Empires in particular, breaking up and spinning off states with somewhat arbitrary borders that are larger than the rulers can actually control but are still treated by the international community as countries. So in some sense in the last 200 years there have been massive state break ups, but compared to say 1700 we have a lot more big states and a lot fewer smaller kingdoms and the like. Things got kind of crazy there with the industrial revolution and all.
In general, however, I lean towards technology being a stabilizing force, working to reduce diversity across groups, bringing them closer together in terms of time and distance, with the resultant cultural differences dropping. Technology mitigates diversity, in other words. In math:
Happiness = -Strictness * (PopulationSize*Area)/Tech
One more thing on Strictness
Coming back to an earlier point before I call this done, remember that Strictness is relative. Very few people genuinely want very few laws; most people want lots of laws provided those laws match very closely what they personally think people ought to be punished for. “There ought to be a law…” is a common phrase for a reason. If your subjects are very homogenous their sense of what the law should be governing with a light touch is likely to make them less happy than enforcing many laws that match their preference. It is only with a population that wants very different laws that the Happiness maximizing strategy is to rule with a light hand.
The question of Strictness is then a question of your deviation in Strictness relative to the variance in the population. A best case policy is to minimize the difference between the demands for strictness from your population and your policies, accounting for the magnitude of those preferences and willingness to accept deviations. Good luck with that!
As a general rule, one must keep an eye out for Taleb type tyrannies of the minority, where a minority cares a lot about a particular rule while the majority is largely indifferent. Taleb gives the example of Kosher orange juice: those who keep Kosher care quire a bit while everyone else is oblivious, the cost of making orange juice Kosher is apparently very low, and so all orange juice in the US is Kosher. Maximizing happiness suggests legislating those sorts of rules when possible.
The trouble occurs when the majority is not indifferent, or when other minorities are deeply antagonistic to those rules. If one tried to ban the consumption of meat in the US, there would be consequences. Probably the majority of Americans don’t eat bacon or sausage for breakfast, but try telling them they can’t and you are going to have a happiness problem.
How would one walk the line between a strident minority demanding people eat less meat and a majority population of omnivores? There probably isn’t a way that would make everyone happy. The best you could probably do is impose policies that subtly raise the price of meat such that people cut back consumption as prices just get too high. You could perhaps do this by subsidizing the production of fuels from animal feed, for instance. You can avoid the stigma of actually putting your name on a law making the activity illegal, yet still market your actions as enforcing the preferences of a particularly strident minority.
So, if we are to make one final change to our vaguely mathematical model it would be adding some minimum or maximum deviation function to Strictness:
Happiness = -dev(Strictness) * (PopulationSize*Area) / Tech
Why Did I Spend Over a Month on This?
Uhm… good question. In part I wanted to build this model so that I could use it to tell a story about the modern USA, and why things seem to be splintering so hard. As it turns out it will also be useful to make predictions about what will happen in Ukraine given different possible outcomes of the war. In general I think it is very useful to understand why nations that seem to be doing alright suddenly start to come apart at the seams, and why some structures of rule work better than others depending on the situation.
At any rate, when can redefine “empire” as:
A state where the Happiness value has reached a low enough state that populations will attempt to split off. These population splits will continue until either effective Diversity decreases, Technology increases, and/or the deviation between Strictness of rule and desired Strictness decreases such that the Happiness value increases to an acceptable threshold.
So long as overall the population is happy enough to want to remain within the state, you have a Great Nation. When the state needs to start actively keeping some of the people in line lest they break away, you have an Empire, possible an Evil Empire.
Will this help you read the news? No. They are going to keep using “empire” to mean “state we don’t like,” regardless of characteristics. It should, however, help you think about how difficult it is to keep ever changing groups of people together under one state, and why things get so complicated.
Whew, yea this one took forever, sorry. Partly I can blame the job search, since interviewing is exhausting, but just as much it was just a slog to write. I have three or four different versions lying around my pc, either in proper folders or the recycle bin, simply because it is an abstract model of an abstract concept, and I kept getting on on tangents and the like. I had to promise myself I wouldn’t write on any other topics until I got it finished, lest I consign hours of work to the void simply because I couldn’t decide whether or not to write a bloody equation down. Thanks for your patience!
Later we might be able to integrate these overlapping power centers into the model.
Sallust apparently said something like this once, I am told.
I am using law here in the colloquial “What the government legislates” sense, in part because I don’t want to type legislation over and over, and don’t want to confuse people.
If you are not so sure that people will demand more strict rules instead of more freedom, spend some time on political forums on both sides of the US party divide. It won’t take long before you come across plenty of folks that that disagree with the other side not because they are taking away their freedom, but because they are mandating the wrong thing. In fact I expect that you will find that far more common than actually complaining about loss of freedom.
Whether these reputations are more or less accurate is another question.
More on technological factors in a bit.
A friend of mine moved to the US from France, and chatting with him at a conference in New Orleans about how southerners are comparatively chatty like that he responded “What are you talking about? All Americans are like that. In France, you could sit next to the same person on the train for hours, and if you start a conversation, you are the weird one.”
For example, see the current state of water rights in California. Then consider how reasonable it is to grow almonds in CA, considering they are very water intensive crops.