Imperial Statistics (NP: Wei through Chen dynasties)

A forum for guides, lessons and sharing of useful information.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Imperial Statistics (NP: Wei through Chen dynasties)

Post by Salmoneus »

My latest half-hearted curiosity: wondering how long Roman Emperors tended to reign.

Why? Well, general curiosity, I guess, but also with a conworlding interest: if you have a polity in some ways similar to the Roman Empire, how long should you allocate to the rule of each monarch? You can have a guess, of course, on the basis of plausible life expectancies and preconceptions about the age at which someone might come to power. But another way is to look at the real Roman Empire for inspiration. So that’s what I did. And whereas in the past I’ve cast a glance over lists of rulers to glean an impressionistic notion of what was commonplace, this time I decided to actually run the numbers. In addition to the length of reign, I also made note of ages at accession and death (or other removal from power), and categorised the methods by which Emperors came to and left the throne.

I don’t know how useful this will be to anyone. In particular, it probably won’t be that meaningful until or unless I run some more numbers on other empires and kingdoms to give some comparative context. But it’s a start, at least, and the results were in many ways no what I was expecting.
I’m going to put the caveats and definitions in a post at the end. But for now, let’s just say: I tried to define a list of Roman Emperors enduring from the rise of Augustus to the death of Constantine IV (considering rulers afte that to be Byzantine rather than Roman). And my findings were...
Last edited by Salmoneus on 03 Oct 2022 15:28, edited 1 time in total.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Across the 712 years in question, the average Emperor achieved a reign of 10 years – taking the throne at the age of 40, and surrendering it, one way or another, at the age of 50.

Reigns varied greatly in length – from less than a year to over 40 (in the case of Theodosius II) – but they were very much not evenly distributed: while 30 of the 70 Emperors reigned for 10 or more years, only 14 made it to 20, and only 4 made it to 30 – 1 of those 4 being Augustus himself, and the other three all reigning after the Sack of Rome. The median Emperor, therefore, reigned for only 7 years (and was 3 years older than the mean).

In terms of the range of ages, meanwhile, a quarter of Emperors ascended between the ages of 7 and 20; another quarter, between the ages of 55 and 76. The youngest age at which an Emperor lost their throne was 7, although in total only four lost power as children or teenagers (Leo II and Heraklonas, who both ruled less than a year, along with Elagabalus and Gordian III). On the other hand, 26 Emperors made it into their 60s, of whom 9 reached their 70s; 2 of those 9 even reached their 80s (Justinian and, oldest of all, Anastasius). No Emperor over the age of 63 was overthrown by a rebellion, and no Emperor over the age of 72 was murdered (that we know of, although of course it can be hard to distinguish murder from natural causes at such ages). Of course, any Emperor in danger of being murdered probably wouldn’t have been in a position to make it to 72...

As regards how Emperors came to power, only 13 were the biological son of their immediate predecessor; a further 24 inherited the throne in some other legal fashion (adoption as a son, specification as preferred heir (including elevation as junior ‘co-Emperor’), inheritence by some other regular relationship (brother, grandson, nephew), or in one unusual case marital selection by an heiress wife). Surprisingly, only 12 came to power through military conquest – that is, at the head of an army marching upon Rome – while 21 were in some way proclaimed Emperor by a third party upon the death of the incumbant (by a general, by a committee of soldiers, or by the Senate).

These routes to power were not equally distributed across time: among the first 35 Emperors, only 3 were biological sons of their predecessor (Titus, Commodus, and Caracalla). In fact, more than half the son-Emperors took power within a period of only 125 years (from Carinus to Theodosius II).
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

As regards departures, half (35) died of natural causes. The remainder were mostly evenly split between those overthrown by armed rebellions (15), and those randomly murdered (15), with the balance, 5, being those who were defeated in battle by a foreign power (3 by the Persians, 2 by the Goths). That doesn’t sound like many, until you realise that it means that nearly 3% of Roman Emperors (within the limits set here) were personally overthrown by Shahanshah Shapur I.


These demises were likewise not evenly distributed: of the first 35 Emperors, 13 were murdered, compared to only 2 of the last 35. Indeed, between the deaths of Probus (killed in protest against swamp-draining duty by his own soldiers in 282) and Constans II (murdered by means of a bucket by his Grand Chamberlain in 668), the Imperial line did not suffer a single, outright cold-blooded murder for nearly 400 years. On the other hand, from the death of Marcus Aurelius in 180 through to the accession of Diocletian in 284, only 4 of the 23 Emperors died of natural causes (Septimius Severus by regular disease, Claudius Gothicus by an unfortunate plague barely more than a year into his reign, Tacitus by a megalomaniacal fever, and Carus by – the ancient accounts implausibly insist – being struck dead by a lightning bolt while ill-advisedly invading Persia).


There are no particularly strong associations between how an Emperor came to power and how he left it. The method of accession is related, however, to their age, and the length of their rule. Those elevated by proclamation were on average oldest, at 48 (nearly 49), reflecting that such men had generally established a reputation over a long career – indeed, this category contains the two oldest men to become Emperor, Galba and Tacitus (both proclaimed by the Senate), although it also includes two teenagers (Elagabalus, elevated as the child figurehead of a rebellion, and Gordian III, elevated by the Senate in a desparate attempt to find someone the masses wouldn’t immediately murder). Unfortunately, these appointees were by far the least succesful Emperors, at least in terms of survival odds, on average ruling a paltry 4 years (neither Gordian nor Elagabalus survived to adulthood). Nonetheless, a few proclaimees did make a mark: Constantine ruled more than 20 years – if he can be counted in this category, it’s complicated (he could arguably fit in any of my categories) – while Claudius, Valentinian and Leo each made it at least a decade.


Biological sons, on the other hand, were by far the youngest to take the purple, averaging only 22; they on average served a perfectly respectable 12 years, but due to their young age this means that they died at the age of only 34. This should also make us perhaps reconsider our idea of what constitutes ‘natural causes’ – note that while the average conqueror left office (i.e. in almost all cases died) at 58 (including all those known to be murdered or executed!), the average son-Emperor believed to have died of ‘natural causes’ died at 37. Only a single son managed to make it to 50! Either the Imperial mantle was exceptionally stressful, or a goodly number of those ‘natural’ deaths must be assumed to have been murders...
The longest reigns tended to belong to those who inherited in other ways (13 years); however, conquerors lived longer (58, to inheritors’ 52) – the difference is because conquerors on average came to the throne a decade later.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

We must recognise, however, that the nature of imperium changed over these 712 years (I’ve given a few examples above). So we can run the numbers on a few subsets of time.

First, if we subtract the Crisis of the Third Century – from the death of Severus Alexander to the accession of Diocletian – we lower the average age of accession to 38, but increase the average reign to 12 years. Surprisingly, this leaves the average age of death the same (in fact, it slightly increases it, though by less than a year), as these two effects cancel out. It does, however, increase the chance of a ‘natural’ death from 50% to 57%, while reducing the chance of being murdered from 21% to 17%, and halving the chance of being killed or enslaved by a foreign power from 7% to 3.5%. Surprisingly, the odds of being overthrown remain the same – surprisingly few Emperors were overthrown in the Crisis, largely because they were murdered before they had that opportunity. During the Crisis, on the other hand, the chance of a natural death crashed to only 21%, the odds of being murdered rose to an intimidating 35%, the odds of being overthrown remained the same as discussed, and the chances of falling prey to Shapur (and/or the Goths) rose to 21%. Despite the era’s reputation for internecine conflict, it was actually the only period in which Emperors were just as likely to fall to a foreign army as to a domestic one (if, again, being murdered by random soldiers doesn’t count). This may be because many Crisis Emperors were naturally military men, or may reflect a need such Emperors saw to quickly build reputations for themselves and rally support around the flag by personally leading armies against the Empire’s enemies.


More generally, we can split the Empire approximately in half, and compare the period from the rise of Augustus to the death of Constantine (364 years, more or less) with the period from the death of Constantine I to the death of Constantine IV (around 348 years). [this is both as close to halfway as you can get if you want an integer number of Emperors in each half, and also of course a hugely symbolic turning point in Roman history; we could of course pick the rise of Constantine instead as our breakpoint, but this is methodologically messier, because he was not initially a sole Emperor]. For the sake of convenience if not 100% accuracy, let’s call these the Pagan Emperors and the Christian Emperors (though “Roman” and “Constantinopolitan” might be more to the point...).

Among the ‘Pagans’, the odds of a natural death were only 38%; those of murder, 33%; those of overthrow and foreign defeat remained the same as for the whole. [so murder wasn’t actually much more likely in the Crisis than at other times in the first half of the Empire]. The average age of accession was 44, and of death (or enslavement, or abdication) was 52. The average reign was only 8 years, and the median reign a paltry 5 years. Proclaimed Emperors only did a little worse than across the whole Empire (3 years rather than 4, but in reality the difference is a matter of months), and non-son inheritors did a little better, at a whole 14 years, nearly 15. But the real catastrophe here is for the firstborn sons, whose reigns lasted, on average, only 6 years, half the length of the overall average. Rebellions were also impatient: the average Emperor overthrown by a rebellion had ruled for only 3 years.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Among the Christians, by contrast, the odds of a theoretically natural death soared to 67% (almost double!), while the odds of (obvious) murder plummeted to only 3%, an astonishing decline. The odds of death in battle against an enemy power, and of being overthrown by rebels, remained (exactly!) the same. However, rebels were more willing to give an Emperor time: the average Emperor who lost their throne to a rebellion now did so after 7 years of rule, more than twice as long as in the Pagan empire.

On the other hand, everyone was now much younger. The average Christian Emperor came to the throne at only 34, a whole decade younger than the Pagan Emperors, and left it at 47 – in other words, the average Christian Emperor died at an age when the average Pagan Emperor was not yet even an Emperor. Emperors died younger, yet they reigned longer – 13 years, on average (the media reign was 11 years, more than twice as long as for the Pagans). Given that obvious murder was far, far, far less common, and that death by rebel or foreigner was no more likely, we are left again with the hypothesis of ‘Imperial Stress’ – a mysterious medical condition that leads Emperors to die abnormally young, yet of ‘natural causes’. Perhaps this may be blamed on poor diet (one Emperor did die of gout), or anxiety-induced heart disease. Perhaps genetics are a factor: notably, while Christian conquerors, proclaimees and non-BioS (biological son) inheritors all on average survived into their 50s (proclaimees lived longest (56), though they still had the shortest reigns (8 years)), BioSs on average died at 33. This is remarkable, given that 75% of them died of ‘natural causes’, and 0% by ‘murder’. Or perhaps the Imperial chefs were just really bad at working out which mushrooms were safe to eat. In any case, it certainly seems as though becoming Emperor was the worst possible thing to do if you cared about living to see old age.

[in general, Roman life expectancy, estimated through multiple methods, seems to have been between about 18 and 25 – astonishingly low, comparable with the worst slums of the Industrial Revolution (life expectancy in Britain in 1840 was around 40, but that of a male factory worker specifically was around 15). This is driven, however, by brutal child mortality – while an 18th century British baby had a 75% chance of surviving to the age of 5, a Roman baby’s odds were more like 60%. 50% of people were dead by the age of 10. More generally, early death no doubt partially explains the fate of BioSs – they came to the throne in the Christian era at the age of only, on average, 17, with several more years of adolescent illness to come. Constantine III, for example, died at 28 of tuberculosis, a disease that famously ravaged Roman men in their 20s. Nonetheless, a Roman boy who survived to adulthood should on average have expected to live to 45 or 50, so the fact that BioS Emperors, with all the advantages of wealth and rank, only survived to their mid-30s is certainly striking.]

On the other hand, because BioSs also came to the throne earlier their average tenure was still impressively long: 16 years (compared to, as mentioned above, 6 years for the same class among the Pagan Emperors). Indeed, other than three who died within five years, all the other Christian BioSs ruled longer than 10 years, including three with reigns of 24, 28 and 42 years (that is, 3 of the 7 longest Imperial reigns).
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Overall, it’s clear that there was over time a transition in the nature of imperium. The process became more regular, with less need of ad hoc proclamations – only 14% of Christian Emperors, compared to 43% of Pagans. Irregular, non-BioS inheritance became more common (46%, up from 26%), but BioS inheritance was the biggest gainer, soaring from 12% to 28%. Conquest also became less common – only 10%, down from 19%. The shift from proclaimed Emperors to BioS Emperors is the main reason for the decline in average age of Emperors. Interestingly, although fewer Emperors rose through conquest, a similar percentage of Emperors continued to be overthrown in rebellions. This implies that rebellions became less likely to place their leader on the throne. It may well be that the shift toward BioSs partially explains this as well: younger Emperors with less experience, knowledge, and personal connexion to other powerful figures were often more attractive to, as it were, benign influencers, making the ‘altruistic’ rebellion more appealing for potential powerbrokers than before.

An archetypal example of this process can be seen in the career of Valentinus: in 641, Valentinus entered Constantinople at the head of an army and deposed the teenage Emperor, Constantine Heraclius, known traditionally as Heraklonas*, exiling the Emperor to Rhodes with his nose cut off (where he conveniently died of unknown causes a few months later). In earlier centuries, this would almost certainly have resulted in Valentinus becoming Emperor... but in 641, it did not. Instead, Valentinus placed Heraklonas’ nephew on the throne as Constans II – at the wise old age of 11. Four years later, Valentinus changed his mind, and attempted to storm the capital once more, demanding that he himself be enthroned... but the public (and the religious and military authorities) rejected this demand, and instead publically murdered him. This would all have been rather surprising a few centuries earlier – many Pagan Emperors had attempted to secure the succession for their children, but the children had generally never received any loyalty and tended to be murdered when the father died. A victorious general would simply have seized power – not only would they not have imagined ruling through a juvenile puppet, but the public would not have imagined preferring that puppet to a seasoned, experienced general.

[it’s also interesting that the first five Roman Emperors under the age of 25 became by-words for depravity and tyranny for millennia to come: Caligula; Nero; Commodus; Caracalla; Elagabalus (the sixth, Severus Alexander, wasn’t a great improvement on that score). Now, to be sure, accounts are biased, as most historians were conservative aristocrats, and reputedly Caligula, Nero, Commodus and Caracalla were all extremely popular rulers with the general public. But nonetheless, nobody denies that they were all in their own ways famous for excess, and for controversial, innovative policies (the same is true of Elagabalus, although his popularity was shortlived). Later young Emperors, on the other hand, tended to be far less controversial (Heraklonas was unpopular not for any known personal sin, but as the stooge of his hated mother, and as an obstacle in the path of the beloved Constans II) – rather suggesting how much less actual power they had than their predecessors. Meanwhile, long-reigning Christian BioSs like Honorius and Theodosius II were famously dominated by their relatives and advisors].

*could we just pause to show our anger toward both Roman parents and later historiographers in the matter of naming? Emperor Heraclius had two imperial sons, whom he helpfully named Constantine Heraclius and Heraclius Constantine. You might think that Constantine Heraclius would be known to history as Constantine III, and Heraclius Constantine would be known as Heraclius II... but no. It’s Heraclius Constantine who is known as Constantine III, and Constantine Heraclius who is known as... Heraklonas. Yes, it’s a nickname. And it has a ‘k’ in it instead of a ‘c’, for no particular reason. Constantine III’s son, meanwhile, was Flavius Heraclius, known to history as... Constans II. Why? Was he called ‘Constans’? No. He was born ‘Flavius Heraclius’, and he chose to reign as ‘Constantine’, but historians prefer to call him ‘Constans II’. For absolutely no reason. But just for fun, he can also be called Constantine III (not IV? No, III, same as his father), or Constans III (because Flavius Heraclius was, confusingly, only the second Constans II). But at least in this case historians might claim they’re just trying to find a path though the confusion. My least favourite ‘historian name’? Philip the Arab. OK, his name was Philippus, although this was a surname, not his personal name (he was Marcus Junius Philippus, to be precise). But why ‘The Arab’? Yes, he was an Arab, but there is no other Emperor Philip to confuse him with! Unless you count his son, ‘Emperor’ Philip the Younger (one of those unfortunate “your father is dead now so we must murder you” kids)... who was also an Arab. It’s like introducing your co-workers as “Bob”, “Joe”, “Mike the African” and “Mike Jr.” – it’s not technically inaccurate, but why, in the modern day and age, do you feel the need to do that? It’s not like if you talk about an Emperor named Philip doing things (other than sitting in a palace playing with toys and then being murdered) anyone’s going to be confused about which one you mean, and if they are you can always specify ‘the Elder’, as you would with anyone non-Arab! Or just, you know, Philip I. *sigh*
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Anyway, we can clearly see a transition – albeit still incomplete – from an early practice of nomination of rulers, either by a preceding ruler or by some other authority, political or military, toward a later practice of just crowning a nearby pliable young relative of the last guy. This is associated with younger rulers, and probably represents a shift toward a sort of constitutional monarchy – authority is invested in the family, not just the individual, but because that authority is not personal in origin the individual often lacks personal power.

The fascinating question, of course, is whether this shift – which does seem to have delivered more stability, with longer average reigns – simply happened too late to save the Empire (these ‘Christian’ Emperors were at best Emperors of the East (excluding Honorius), and over the course of this period lost almost all of the East as well), or whether to some extent this shift was actively promoted by the shift toward a smaller, more manageable Empire with more centralisation of power. [it’s also possible it represents a more specific cultural trend – not that the Latin-speaking Emperors avoided terms associated with monarchy, while the Greek-speaking Emperors happily called themselves ‘kings’, perhaps because the hostility toward kings present in Rome had never been a significant force in Greek culture].

We can also, incidentally, look at this process from the other direction, and ask about the ‘success’ of Emperors in elevating their own biological sons or nephews. For this measure, I’ve not only counted immediate succession – a son or nephew following the father – but all Emperors whose son or nephew later became Emperor. In the Pagan Empire, only 14% of Emperors were ‘successful’ in this way (though the much larger number of Emperors who at some point attempted to secure the succession for such a close relative indicates that this was indeed a common objective for them). In the Christian Empire, by contrast, 50% were by this measure ‘successful’.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

So, there we go. What does any of this mean? I don’t know. I think it will be interesting to compare these stats to those of other Empires, or those of mediaeval kingdoms (including Byzantium). For now, I’ll recap some headlines.

I’ve introduced here another measure here that I hope may be useful: quartile reigns. For this, I've given three numbers - 1/4 of Emperors ruled for less than the first number of years; 1/4, for between the first and second number of years; 1/4, for between the second and third number; and 1/4 for more than the third number. I've given a third number in brackets - this is the mean age of Emperors in the upper quartile. Thus, for example: in the 'Early Roman Empire', 1/4 of Emperors ruled less than 1 year; 1/4 ruled between 1 and 5 years; 1/4 ruled between 5 and 13 years; and 1/4 ruled more than 13 years, among whom the average reign was 20 years.

So:

Overall: accession at 40; 10 years of rule; death at 50.

Early Roman Empire (Augustus to Constantine):
Mean Age at Accession: 44
Mean Age of "Natural" Death: 64
Mean reign: 8 years
Quartile reigns in years: 1, 5, 13 (20)
Death: 52
Immediate succession by biological son of predecessor: 12%
Eventual succession by biological son or nephew: 14%
Survival to a theoretically “natural” death: 38%
Most common route to throne: proclamation by third party (43%)


Late Roman Empire (Constantine I to Constantine IV):

Mean Age at Accession: 34
Mean Age of "Natural" Death: 52
Mean reign: 13 years
Quartile reigns in years: 3, 11, 21 (30)
Death: 47
Succession by biological son of predecessor: 28%
Eventual succession by biological son or nephew: 50%
Survival to a theoretically “natural” death: 67%
Most common route to throne: inheritance not from biological father (46%)


Hope something here will be interesting or useful to somebody!
Last edited by Salmoneus on 08 Sep 2021 16:55, edited 1 time in total.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Finally, I’m afraid I feel the need to run through the caveats pedantically. So here we go:
Spoiler:
First, the time period. I decided to end with Constantine IV, for three reasons: by the end of Constantine’s reign, it was clear that the Roman system had been destroyed, following catastrophic territorial losses, massive de-urbanisation (Constantinople lost 9/10th of its population) and the creation of the thematic system, effectively bringing about a transition from the Eastern Roman Empire (an empire) to the Byzantine Empire (really just a local kingdom, albeit an important one) – this obviously isn’t an objective clear line and could probably be moved one or two Emperors earlier or later, but Constantine IV seems like a good watershed; choosing Constantine IV as the last Emperor gives me a nice round number of Emperors; and this also lets me avoid having to decide how to deal with Justinian II, who was Emperor on two non-consecutive occasions. Meanwhile, in the West, I’ve only gone as far as Honorius; subsequent rulers were controversial (most weren’t acknowledged by the East), ruled only a small area, and in any case were clearly in no way the equal of the Emperor in Constantinople.

Next, I attempted to count only one Emperor at a time where possible - that is, I've tried to count people who actually ruled the empire, rather than everyone given a courtesy title. This means that, for a start, I haven’t counted any of the many child ‘co-Emperors’ who were given official rank as a way of indicating that they were the chosen heir, but who didn’t actually hold any power (and who in almost all cases were murdered before reaching adulthood anyway). That also goes for a couple of adult-heir co-Emperors (hello, Lucius Verus). I’ve also discounted junior Emperors who remain clearly subordinate to their ‘colleague’ – so Maximian, for example, is not counted, as I’ve considered him clearly subordinate to Diocletian (Diocletian was, for example, able to compel Maximian to abdicate, twice). I’ve only considered there to be two Emperors at once when neither clearly dominates the other. For the record, this means Licinius/Constantine, Constantine II/Constantius II, Constans/Constantius II, and Honorius/Arcadius.

Then there’s the question of when someone becomes Emperor at all. I’ve decided not to count everyone who declared themselves, or was declared, Emperor – traditional lists often seem to make completely arbitrary decisions over which rebels to include, and I’ve just not included any of them. In my list, you’re not an Emperor until you have clearly defeated the incumbant (and yes, this means I have Gordian III, but neither Gordian nor Gordian II – sorry, but you’re not The Roman Emperor when your sole claim to fame is a couple of months of semi-authority over parts of Tunisia, I’m just not having it). To make it clear that you’ve actually become Emperor – rather than happening to be in Rome while a civil war is going on – I’ve also insisted that a candidate actually remain ‘Emperor’ for at least three months. This also helps deal with some annoying little “who died first?” questions here and there. So, no Pupienus and Balbinus, and no Didius or Nymphidius, sorry. If someone declared you Emperor but then everyone deserted you a week later when the real Emperor turns up to kill you as a rebel, you weren’t really the Emperor.

Next, there’s the cases where somebody rebelled, and gained effective autonomy for a prolonged period of time, but never actually overthrew the preceding Emperor, and only maintained rule over a clearly smaller territory than the actual Emperor. I have not counted these people as Emperors – so I’ve ignored the Gallic and Palmyrene “Empires” and the various regional claimants during the Wars of the Tetrarchy. As a borderline case, I’ve also excluded Magnus Maximus – although he controlled most of the West, he remained clearly inferior to Theodosius, who eventually defeated and killed him.

Finally, where a rebel gains temporary control in the course of an ongoing civil war that they ultimate lose, I’ve not considered the rebel to be an Emperor. This means in particular that I haven’t included Basiliscus – yes, he controlled Constantinople and virtually the entire Empire for a year, but Zeno was still alive, did not concede (he was beseiged in a fortress), retained control of the Imperial treasury (which he stole when he fled Constantinople), and ultimately returned to power and immured Basiliscus in a cistern. I’ve considered this to simply be an ongoing civil war, rather than a clearcut transition of power to Basiliscus, even though Basiliscus had an overwhelming military advantage for a brief period of time.

Having decided who was and wasn’t Emperor, there’s then the question of ages and reigns. For simplicity, I’ve rounded everything down to the nearest year. If you’re 59 and 8 months, you’re 59. If you ruled 1 year and 8 months, you ruled 1 year. Yes, this means some people ruled 0 years.

Next, there’s classification. Obviously, there’s a lot of ambiguity in some cases about how someone came to power. There’s probably some listed here as inheritors who weren’t really – certainly that’s been claimed for some of them. It’s unclear, for instance, whether Trajan ever ‘officially’ adopted Hadrian, or only considered him a potential future adoptee. I’ve simply gone with the official story that was widely accepted, as this at least reflects how people saw the Emperor’s source of authority: if people were happy to say that Hadrian was Trajan’s heir, I’ll go along with it – that’s the basis on which Hadrian claimed power, whether or not it’s factually accurate.

Then there’s deaths. As discussed, a bunch of them are probably murders, but the problem is, almost every notable historical death was considered murder by somebody – “our beloved leader was secretly poisoned by the Evil Ones, making their rule illegitimate!” is the oldest and most perennial conspiracy theory in politics. I’m not a historian, and even if I were the truth in most of these cases is unknowable. So I’ve simply assumed that people weren’t murdered, unless it was generally accepted that they were – I’ve given their ‘poisoners’ the benefit of the doubt where possible. And yes, Jovian, that includes you, even though “our unpopular Emperor, shortly after the humiliating defeat, was mysteriously poisoned by accidentally inhaling toxic fumes from the fresh paint on a nearby wall unwisely near to which a brazier had been placed overnight, how sad” does read like the most pathetic defence case ever offered. The fact is, freak accidents do happen, there’s no clear evidence that Valentinian had him killed, and there was no general contemporary belief that he’d been murdered, though some did suggest it. And talking of freak accidents, let’s address the elephant in the room: yes, I’m accepting that Carus really was struck by lightning. It’s so fantastically unlikely that a lot of historians have assumed it must be propaganda, but there’s no actual evidence for this that I can see beyond plain skepticism – and indeed, there’s evidence that his army were indeed really freaked out about something, which would make sense if their Emperor had just been struck dead by Jove himself. So, I’m accepting it.

And then there’s poor Gordian III. Nobody knows what the fuck happened to him. Walked into a desert with his army, his army walked back out, he did not. The Roman explanation – everything is fine, nothing happened, we didn’t lose a battle, by the way we have a new Emperor now and a humiliating treaty with the Persians – is obviously nonsense, so I’ve just accepted the Persian explanation (Shapur – or at least his men – killed him). It’s entirely possible, though, that he survived the battle but was murdered in a conspiracy by Valentinian, or in an unplanned mutiny by despairing soldiers, or indeed that he did just die. We don’t know, and can’t know, but we do know he died in very close conjunction to a probable (if unclear) battle, so I’ve counted him as killed in battle.

Speaking of which: I’ve taken a simple approach to Emperors overthrown by rebels, counting them all in one category. In reality, some were killed in battle, some were murdered by their own men immediately after a battle out of anger at their failed leader or fear of his replacement, or before a battle, or by the winner of a battle, while others comitted suicide either at the command of the victor, or out of genuine despair. Sometimes it’s unclear exactly what happened and why. I’ve just gone for simplicity, and said that anyone who dies in the immediate context of an armed rebellion against them, the leader of which then becomes their successor, is assumed to have been overthrown by those rebels, regardless of who actually held the knife. Conversely, I’ve only counted them as outright murdered when the murder does NOT happen in this context.

We’re also left with two outright anomolies: Valerian, who was enslaved by Shapur (later Christian sources insist he was flayed alive, but it’s just as likely that he remained a political hostage until dying of natural causes); and Diocletian, who abdicated absolute ultimate power and unimaginable wealth in order to have more time for his primary interest, cabbage farming. [during the ensuing civil war, people begged him to come back, but he insisted that the cabbages were much too important for him to waste time on all that god-king business]. For the sake of simplicity, I’ve lumped Valerian in with the Emperors killed in battle against a foreign power, and I’ve considered Diocletian to have died of natural causes (which he probably did, eventually, at some unknown point).
Hope that's wrapped up any loose ends!
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Roman Emperor Statistics

Post by Creyeditor »

Really fascinating and helpful. I always thought an average reign of 5 years (in conworlding) was too long, but it seems that's not the case. Also, interesting too see that non-institutionalized methods of succession were somehow more common and more stable. Did I understand that correctly?
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

The weird thing about the Roman Empire is, ALL succession was non-institutionalized. There was never any organised way to select a new ruler, or any agreement so far as I can see as to what would constitute legitimacy per se. Or rather: the theoretical method of becoming Emperor, election by the Senate, was at best generally a rubber stamp, and frequently given no consideration at all. It very rarely actually operated independently, and when it did so it was not successful. Essentially, I think this is because the Senate never really had much legitimacy in its own right.

So, I've kind of imposed a framework in which primogeniture is ordinary and institutionalised - because that was the norm in later western societies - but for the Romans, at least in the early empire, it was only one of several possible arguments for an individual being given the throne. It was rarely couched in terms of an institutional right to rule, but rather in terms of loyalty to the parent, or expectation that the parent's virtues would be passed on to their son.

But yes, biological primogeniture always accounted for only a small minority of imperial successions - although it did increase noticeably in frequency in the latter half of the empire.

In terms of stability, that's one of the weirdest things. In the early empire, biological sons were among the shortest-reigning emperors (and often among the most controversial), reigning only 5 years (I misspoke when I said 6 before), and that's discounting all the ones that were put to death almost immediately. But in the later empire, they were the longest-reigning emperors, reigning 16 years.

I think this is because Roman authority was personal, and as individuals biological heirs are generally bad candidates for rule: they lack personal accomplishments, they're often spoiled little brats, and they're typically very young. Most potential biological heirs struggled to be accepted - most, indeed, were immediately rejected. But over time, authority began to be more institutionalised.

This is vividly displayed in ages - very few people in the early empire came to the throne as teenagers, and none of them ended happily (in fact, nobody who came to the throne at the age of less than 39 managed to die of natural causes). But in the later empire, you don't just get teenagers, you get Emperors who are 10 or younger. While a teenager like Caligula could sometimes rule at least in part on their own authority, a 7-year-old is clearly not being honoured as an individual, but as an institution.

Essentially, the more you move away from a personalised, meritocratic system of rule toward a system of arbitrary genetic inheritence and divine right, the more stability you get!
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Incidentally, I'm looking at English monarchs now, and they may as well be in an entirely different universe...
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

OK, here are the English headline stats, from Alfred through to Anne (she seemed a good stopping point; not only do they stop being officially 'King of England' at that point, but they also become almost entirely constitutional monarchs):

Mean Age at Accession: 27
Mean Age of "Natural" Death: 47
Mean reign: 17 years
Quartile reigns in years: 5, 13, 23 (37)
Immediate succession by biological child of predecessor: 40%
Eventual succession by biological child, niece or nephew: 64%
Survival to a theoretically “natural” death: 71% [slightly distorted by counting Edward IV and Henry VI as two people each...]
Most common route to throne: direct succession (40%)


We can divide this up into eras:

Saxons and Danes (Alfred to Harold II):
Mean Age at Accession: 24
Mean Age of "Natural" Death: 37
Mean reign: 11 years
Quartile reigns in years: 3, 9, 17 (30)
Immediate succession by biological child of predecessor: 33%
Eventual succession by biological child, niece or nephew: 67%
Survival to a theoretically “natural” death: 86%
Most common route to throne: succession by biological child of previous monarch (47%)

Normans and Plantagenets (William I to Richard III):
Mean Age at Accession: 26
Mean Age of "Natural" Death: 55
Mean reign: 21 years
Quartile reigns in years: 10, 18, 34 (44)
Immediate succession by biological child of predecessor: 47%
Eventual succession by biological child, niece or nephew: 74%
Survival to a theoretically “natural” death: 58%
Most common route to throne: direct succession (47%)

Tudors and Stuarts
Mean Age at Accession: 32
Mean Age of "Natural" Death: 50
Mean reign: 16 years
Quartile reigns in years: 5, 13, 23 (35)
Immediate succession by biological child of predecessor: 38%
Eventual succession by biological child, niece or nephew: 46%
Survival to a theoretically “natural” death: 77%
Most common route to throne: direct succession (38%)
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Roman Emperor Statistics

Post by Creyeditor »

What is the difference between immediate, direct, and eventual succession?
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Creyeditor wrote: 08 Sep 2021 22:21 What is the difference between immediate, direct, and eventual succession?
Sorry!
I was using 'immediate' and 'eventual' in contrast: immediate succession relating to A being the first successor of B, and eventual succession relating to A being a ruler at some point after B.

I've used this is two different measures: immediate succession by a biological child (as a fact about how the child comes to the throne), and eventual succession by a child or nephew (as a fact about the dynasty-building success of the parent/uncle/aunt). [the latter seems a bit random - maybe it would be more meaningful if I threw in siblings as well?]

I've used 'direct succession' to mean smply immediate succession by a biological child.
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Roman Emperor Statistics

Post by Creyeditor »

Ok, got it. All in all, it seems to be very different. Succession is much more based on biological relations and the other numbers do not really seem to correlate.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Backstroke_Italics
hieroglyphic
hieroglyphic
Posts: 58
Joined: 27 Nov 2019 19:48

Re: Roman Emperor Statistics

Post by Backstroke_Italics »

I was surprised by these results. My understanding was that the western empire collapsed in large part because they couldn't establish a coherent method of succession, and so they were in a near-constant state of civil war. But this data suggests that the process of succession became more streamlined and peaceful in the later empire.
I did it. I made the world's worst book blog.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Roman Emperor Statistics

Post by Salmoneus »

Backstroke_Italics wrote: 04 Oct 2021 03:59 I was surprised by these results. My understanding was that the western empire collapsed in large part because they couldn't establish a coherent method of succession, and so they were in a near-constant state of civil war. But this data suggests that the process of succession became more streamlined and peaceful in the later empire.
I'd say the inability to establish a coherent method of succession was certainly a huge problem for the Empire, which definitely contributed to the structural decay of the West. This problem came close to destroying the Empire in the Crisis of the Third Century. But they were able to pull back from the brink. Diocletian established a system that was still desperately flawed, but was at least a system - but whether that was the key, or whether broader cultural issues were at play*, I don't know. In any case, it only bought the West another century or so, before it had to be abandoned. The key problem with the West was always that outside of Italy it was massively underpopulated compared to the East, which made it tempting to immigrants and difficult to defend against them - and without the rest of the West, Italy itself was always expensive to try to defend by sea from the East.

There's a considerable streamlining AFTER the fall of the West, but whether that was just too late to the save the West, or whether it was only made possible by the withdrawal from the West, I don't know.


*the fundamental problem I think is that the military by the 3rd century came from an entirely different population from the ruling class, so had no loyalty to it. The military itself apparently recognised this problem - it's said that everyone was really embarrassed after the death of Probus, who was actually quite popular and capable. Maybe cultural change in the military just recognised that instability wasn't a good thing. Or maybe it was to do with the Edict of Caracalla (after a generation to take effect), expanding citizenship, making people feel more loyalty. Or maybe it was just that by Diocletian the old ruling class had largely been usurped anyway. Hard to say...
clawgrip
MVP
MVP
Posts: 2257
Joined: 24 Jun 2012 07:33
Location: Tokyo

Re: Roman Emperor Statistics

Post by clawgrip »

I can't contribute anything meaningful to this discussion, but I just want to say good work on this. Very interesting.
User avatar
Torco
sinic
sinic
Posts: 303
Joined: 14 Oct 2010 08:36

Re: Roman Emperor Statistics

Post by Torco »

cool analysis, Sal.

with regards to the Romans, it's possible that not having an institutionalized method of succession was, rather than a problem, a strength of their system: it might not be an accident that their best years* were the ones with what we might call free or undetermined succession, up to the politics of the time, and these days only shitty states have actual hereditary rule (the ones that are nicer have the trappings, but the monarch isn't that powerful). this might be a bias of myself as not-a-subject-of-a-king (though maybe I am, I never did follow up on that request for Spanish citizenship) but it seems to me that sons of kings are categorically not the best people to be kings, and there's certainly plenty of examples of totally unfit people coming to power by heredity, and Confucius was probably not wrong in saying that shitty rulers are bad for the country they rule. by contrast, succession by politics is more complicated but is more pluriarchical.

* 'best' as in most effective, the roman empire was by all accounts a humanitarian catastrophe.
Post Reply