Quick Take: Roman Weighted Voting & Other Ways to Save Democracy

Several systems beyond one man, one vote have been proposed, and weighted voting was pretty effective in the Roman Republic

There are two reasons why most people go through life having consumed a ton of content about ancient Rome (movies, books, discussions…) and still never get a clear explanation of how the Roman voting system worked: first, it was very complex; second, it was very different, and arguably more efficient, than the one-man-one-vote system used by modern democracies.

Rome worked under republican institutions from around 509 BC to the late 1st century BC. That’s about five centuries of institutional evolution that muddle the picture: any description of the system is going to be inexact for the whole period, unless it focuses on a very specific point of time, and even then the sources are not complete or precise enough to state anything with absolute certainty, in a way that is applicable to all cases.

The reason why these institutions are rarely, if ever, described as a “democracy” is that they weren’t. Using the simplified monarchy-oligarchy-democracy triad popularized by Aristotle, Rome was closer to an oligarchy than it was to a democracy; and this Aristotle would have seen favorably, since he loathed democracy and considered that the worst possible system.

Rome was also, by all accounts, more efficient than any democracy: Athens’ own democracy was created around the same time as the Roman republic, in a blur of murder and trickery, and survived for around a century. Later attempts to bring it back to life were amusingly catastrophic.

The Roman system would be best described, in modern terms, as weighted voting: to simplify, the Roman assemblies divided people by tribal affiliation and social class, and convened in blocs, often but not always centuries. There, a plurality of voters decided each bloc’s position on the issue being discussed (a law, the election of a consul, whatever), much like states are won in US presidential elections.

This, particularly in ancient times, worked great to dilute the effect of populists and outright bribe aficionados like Athens’ Pericles, who were used to sway the large one-man-one-vote Athenian assemblies by making wild promises while keeping their paid, armed mobs at the ready just in case (this may remind you of some modern democracies, perhaps). In Rome, various laws were enacted to make it harder for populists to sway various blocs (“canvassing”).

The system was made more oligarchic and conservative by the fact that Roman tribes were not all equal: the descendants of all-time settlers were in smallish tribes while newer Romans, sons of migrants or former slaves, were attached to huge tribes with the same number of tribal votes as the tiny units where important people voted: one. The US Senate keeps a feature of this system, so freaking Rhode Island has two senators, same as California, the state that is bigger and wealthier than most countries.

Beyond this simplified explanation, there were many bells and whistles and exceptions, and the Republic of the 5th century BC certainly resembled that of four centuries later only in a general sense. It’s also worth noting that the Roman republic worked, as far as we know, in a manner remarkably similar to that of Carthage. I’ve written at length about Carthage’s system here, and about Rome's here, here and especially here, where I describe Rome’s early republic in painstaking detail:

Consuls, originally elected by acclamation, had a great deal of power, but for one year only, and were subjected to prosecution afterwards. The incentive was for them to work within the system and with his peers, thus ensuring a smooth transition to ex-consulship, as long-term influence in Roman politics was often achieved by securing power and support in the senate, rather than any specific measures taken as consul. The senate’s position in Roman society was so elevated, that the august body was only allowed to meet inside of temples.

In time, ex-consuls (known as “consulares”) became the most influential members of the senate, and they were all exclusively drawn from the elite, because no aspiring Pericles was ever able to set salaries for government positions, which stopped anyone poor from even considering public office as a possibility. From 471 BC, the “acclamation” system was institutionalized by making two of the three popular assemblies in the republic – the plebeian council and the military-style comitia centuriata, but not the comitia populi tributa or tribal assembly, in charge of passing laws – vote by “centuries” within the city’s tribes, with wealthier tribes voting first.

This was a very carefully designed system: voting was stopped when both candidates gained the votes of 97 of the 193 centuries (each of which proposed two candidates), so very often many of the poorest, largest tribes and centuries – frequently surpassing the 100 theoretical members of a centuria by far – didn’t even vote. In addition, centuries were split in junior and senior classes, with senior centuries including men over 46, who were significantly fewer than younger ones, but still carried equal voting power.

Such arrangements were complemented by a custom, probably from the 4^th Century BC, to enroll freedmen into the four “urban” tribes of Rome, to minimize their political weight. This was important for the votes in the plebeian council and the comitia centuriata, which elected consuls, but was fundamental for the tribal assembly, where votes were cast by tribe – a step that favored the then-sixteen “rural” tribes, controlled by the patricians to the extent that they were named after patrician families.

Great, but nobody in his right mind would propose the Roman Republic as an example to follow for modern states, right?

Well, no, let me stay on the issue of weighted voting, because I think the Romans were onto something. The way they saw it, they had to give extra weight to the vote of some citizens, because they were, in their opinion, of higher quality: descendants of the first Romans, the people the framers of the Roman constitution wanted to protect and preserve; the Roman constitution says it clearly:

in Order to form a more perfect Union, establish Justice, insure domestic Tranquility, provide for the common defense, promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity, do ordain and establish this Constitution…

Sorry! That’s the American Constitution. But I think you catch my drift. Now, we modern people don’t give hereditary rights so much relevance, but the points the Romans were making are still solid: some people care more, and have more skin in the political game. That’s a fact.

We all have this friend who agonizes over who to give his precious vote to, and looks at the party manifestos and the like. There are people out there like Scott Alexander who actually care a lot about California’s elections, enough to devote time to think who would be the best State Insurance Commissioner, and then write about the issue and explain his rationale.

On the other hand, we all know there’s not only a lot of disinterested and, sometimes but not always, dumb people who vote, whose vote has the same value as Scott Alexander’s, whose rationale to vote goes like this:

-Hey, who are you planning to vote for?

-I love candidate X, he was in the military, he looks like such a straight arrow.

-He wasn’t in the military, and look at all these Internet links about his corrupt dealings.

-Well, they’re all crooks and liars anyway, right?

These people actually form blocs in modern politics. They are constituencies. These people, Pericles’ core voters, chase shiny toys.

These people, not knowing or caring much about actual governance and being easily districted from their core concerns, easily fall prey to single-issue drives and their promoters, as

Curtis Yarvin

keeps arguing when he defends some sort of monarchy. American liberals are thinking of the NRA, conservatives of PETA or George Soros’ machinations.

Several thinkers have argued that welfare programs tend to fatten such constituencies, by giving them a shiny new toy that is actually beneficial to them, and that democracies may be destroyed by the self-reinforcing loop created by proposing better entitlements to cover for the fact that earlier entitlements didn’t end the voters’ misery.

Slavoj Žižek

did it in his book “Lenin: The Day After the Revolution” (2017), warning (presciently, it turned out) that “the reaction to the inability of the welfare state to deliver will be rightist populism.” In Europe we have great examples of this, nowhere more evident than in pensioners’ voting blocs: they love their rising pensions, and if countries are bankrupt after they die, well, that’s just sad.

What I’m saying is that we would all like it more if people like Scott Alexander had a bigger say in politics than just a mere vote. That’s possible, and many people have been looking into historical examples to make it happen.

A modern, more palatable variation of Roman weighted voting is described in “Aurora Rising,” one of the novels of Alastair Reynolds’ masterful “Revelation Space” cycle. In the novel, weighted voting is used in the Yellowstone solar system, inhabited in the 25th century by around 100 million humans, many of whom lived not in the system’s main planet, but in the Glitter Band formed by multiple orbital habitats.

These habitats have many different characteristics, but they’re all subjected to a centralized voting arrangement for important matters, under which each inhabitant has the right to vote: but not with one vote necessarily, but rather through a weighted voted system that provides some with as much as two votes, and many with just over one vote each.

In summary, the system relies on AI to assess the “wisdom” of each voter in each voting round: having identified those voters as being of shrewd judgement, it attaches a weighting bias to any future votes the voters might cast. Most remarkably, Reynolds comes up with the idea that particularly effective voters will tend to gather together in specific habitats (even setting a minimum voting ranking of, say, 1.5 for prospective immigrants from elsewhere in the system) where some are “triples” who have earned three votes after a lifetime of good decisions, and some are incredibly wise “quadruples.”

This is no mere snobbery or selection bias. The citizens of one such habitat in the system make their living by trading on their prior shrewdness: because their votes are disproportionately effective, they are very attractive to lobbyists from other communities. On marginal issues, they pay that specific habitat to listen to what they have to say, knowing that a block vote from such people may swing the result by a critical factor.

Of course, Alastair Reynold’s idea has a catch: you need a potent enough AI to look at voting patterns and the history of various proposals, to judge just how well the humans did. I’m the optimistic type and I actually think that’s a feature, not a bug.

After all, we were promised AI that would clean the bathroom so we would have time to produce art, and instead we got AI that does art, forcing artists into the bathroom-cleaning business. How long before some smarty pants decide that an AI will make better political decisions than either Joe Biden or Donald Trump? With Reynold’s approach we may have the best of two worlds: a smart enough AI that looks at and assesses policies, and humans who actually have the right to decide on those policies, and are not idiots, because they have weighted voting super-powers. It would be a pretty glorious return of Roman-style politics.

There are other proposals in this vein, and I’m sure my readers will come up with a few extra in the comments. I’m thinking of, for example, “delegate voting” or “proxy voting” in the corporate style, where you don’t actually care so much about Tesla governance that you vote in each annual general meeting: you delegate your voting rights to Elon Musk or whoever. This can be done on a thematic basis too, so you delegate health policy to somebody and national security votes to somebody else.

I quite like “storable votes,” a proposal from the Effective Altruism community that does a very good job of replicating the loss of social status and/or power that people who consistently act selfishly or take wrong decisions suffer in small, hunter-gatherer communities and the like. As

Arturo Macias

, a Spanish economist, explains:

At the beginning, all agents are given an equal number of (infinitely divisible) storable votes. The agents say how many votes they are willing “to pay” for each of the possible alternatives and the most voted alternative wins the election. Then, the votes that have been committed to the winning alternative are deducted from each player's account, and are equally redistributed among all participants, and a new voting period begins.

The system reduces the incentives for strategic voting: agents do not stop signaling their interest in alternatives with little probability of victory (if it does not win, you do not pay votes), and it solves the problem of minority disenfranchisement: the more elections a subject loses, the more power future electoral power she accumulated. The article uses exact computational methods (GAMBIT is used for backward induction). The simulations indicate that the PAYW part improves a fixed number of votes version of the Storable Votes  

SV-PAYW shall be considered as a natural alternative to Quadratic Voting for its use in distributed governance systems (vg. to implement the democratic reforms proposed in “Radical Markets”). In my view is equally simple, and the avoidance of strategic behavior is likely to be more complete. Additionally, the sock puppet problem does not exist in SV-PAYW, because the system is linear and “dividing” votes to more electors does not affect electoral power.

I left my favorite electoral reform for last, however. It’s also one that can be said to replicate the best features of early human hunter-gatherer communities. Imagine you are in the forest, and are a young dude; you have an opinion, and an older guy, not even that much older, but one who is already married and has kids, disagrees with your opinion.

In a vacuum, ceteris paribus, whose opinion should have more weight? That of the guy who has nothing to lose, or the guy who has an entire family that depends on his good decision-making? So I’m all for Demeny voting, a concept defined by Hungarian-American demographer Paul Demeny, that would ensure the (indirect) representation of children who are considered too young to vote.

Under a Demeny voting system, parents cast a proxy vote for their child, or even better: half a vote for each, since the parents' political views may differ. This would be, of all the innovations I discussed above, the easiest and fastest to implement, and is one that would make a huge difference, in many senses, in many electoral processes. Let those who have a stake in the future decide the shape of such future, we Romans say.