Every Roman magistrate was elected by the citizens, and the same few dozen families held the consulship for centuries, so the interesting question is how both of those things were true at once.
The previous lesson ended with a puzzle. After 287 BC any citizen could in principle hold any office, and the assemblies that conferred those offices were open to every citizen. Yet the fasti show a narrow and hereditary consulship. Part of the answer is money and prestige, which are true of every aristocracy. The specifically Roman part is in the mechanics of voting, and the mechanics are unusual enough that they have to be set out before anything about Roman politics makes sense.
Nobody voted as an individual
A Roman never cast a vote that was counted alongside other citizens' votes. He voted inside a group, the group produced one verdict by an internal majority, and the groups were counted. Whether your vote mattered depended entirely on which group you were in and how large it was.
The comitia centuriata, which elected consuls, praetors and censors, sorted citizens into 193 centuries by wealth. The richest were counted first, and the count stopped as soon as one side had a majority, because the outcome was then settled and further voting was pointless.
Now do the arithmetic on the older scheme, which Livy describes. Eighteen centuries of equites, the wealthiest men of all, voted first, followed by the eighty centuries of the first property class. That is 98 centuries out of 193, and a majority is 97. If the rich agreed, the election was decided before the second class was called, and the fifth class, the smallest holders, whose 30 centuries contained by far the largest number of human beings, could vote only when the rich were split.
A reform in the third century BC tied the centuries to the tribes and cut the first class to 70 centuries. With the 18 equestrian centuries that gives 88, nine short of 97, so the top of the system now had to carry the second class as well. The details of the reformed scheme are disputed and the direction is not: the change made the richest citizens negotiate with the next group down, and it left the poorest exactly where they were.
Example. Under the older scheme, what did the arrangement mean for a citizen in the fifth property class?
That his vote existed and was almost never reached. He belonged to one of thirty centuries called after 98 others, in a count that stopped at 97, so he was polled only when the wealthy divided. The system was not a fraud, because when the rich did split, the lower classes decided the outcome, and Roman politicians behaved as though they might. But its normal operation was a weighted ballot in which about half of the voting units belonged to a small minority of citizens, and the weighting was explicitly by property, since the classes were originally the ranks in which men were expected to equip themselves for war.
Now you. After the reform the first class held 70 centuries and the equites 18. How many more centuries did they need for a majority of 193, and where did those have to come from?
Answer
They held 88 and needed 97, so nine more, which had to come from the second class or below. The practical effect is a shift in who has to be persuaded: an alliance of the very rich is no longer self-sufficient, and a candidate must reach further down the property ladder. It is a real change and a small one. The reform moved the threshold of decision from the top class to the top two, in an assembly where the poorest citizens still voted last, in the smallest number of effective units, and usually after the result was known.
The tribes, and where the city was put
Laws and the lesser magistracies went through the assemblies organised by tribe, of which there were 35: four urban and thirty-one rural. A tribe was not a place of residence but a registration, inherited from a father, and each tribe cast one vote decided by a majority of whichever of its members turned up.
Two consequences follow, and they run in opposite directions.
First, the city's poor were penned. Freedmen and the propertyless of Rome were registered in the four urban tribes, an arrangement fought over repeatedly in the fourth and third centuries BC and settled by 220 BC. However many of them there were, they controlled 4 votes out of 35, which is 11.4 per cent, against the 18 needed for a majority.
Second, and less often noticed, the rural tribes were cheap. Voting happened only in Rome, on days announced in advance, with no absentee ballot. A rural tribe's verdict was decided by those of its members who were standing in the enclosure that day, which might be a few dozen men, and a great house with clients registered in that tribe could deliver it at low cost. Both effects push the same way: the assembly was a machine that converted a small, organised, well-connected turnout into decisions.
Who was actually in the enclosure
The Roman citizen body was enormous and the space was not. The census of 70 BC registered about 910,000 citizens. Voting took place in the Forum or on the Campus Martius, and the enclosures used for it could hold, on the most generous estimates, some tens of thousands of men, so even a full and orderly assembly would have contained around 3 per cent of the electorate, and the ordinary attendance was far below that.
Distance did the rest. Roman citizenship by the first century BC extended the length of Italy, and a citizen registered near Mediolanum was about 500 km from the voting enclosure, which at a marching pace of 30 km a day is nearly seventeen days each way. A man walking for a month to cast one thirty-fifth of a vote is not a normal event, and it did not normally happen.
This has produced a genuine scholarly quarrel. Fergus Millar argued that the assemblies were real political theatre, where speeches changed outcomes and the crowd's verdict was sovereign in fact as well as in law. Henrik Mouritsen replied that the physical spaces were too small and the electorate too dispersed for anything but a tiny, largely urban, and largely mobilised sample to be present. The evidence supports a compromise that is uncomfortable for both: the votes were real and could go against the Senate, and the body casting them was a small fraction of the citizens, assembled by the people who had reasons to assemble it.
The canvass
What did a candidate actually do? A remarkable document survives to answer that: the Commentariolum Petitionis, a handbook on electioneering addressed to Cicero for his consular campaign of 64 BC and attributed to his brother Quintus. It is practical to the point of cynicism, and it is the best evidence we have for what winning required.
Its advice is to be seen constantly with a crowd, to learn and use names (a slave, the nomenclator, was kept to whisper them), to attach the men who control blocks of votes rather than the voters themselves, to remind everyone of favours done, and above all to promise. A refusal offends immediately and certainly; a promise may never have to be honoured, and if it is broken the injury falls later and on one man.
Behind the handbook lies the structure that made it work: clientela, the standing relation of obligation between a great man and the many lesser men who attended his morning reception, took his advice in lawsuits, and were expected to support him. Nothing about this was secret or shameful, and no Roman thought it corrupt.
Outright bribery was different, illegal and constant. The law was tightened again and again, in 67 and 63 BC among others, which is the usual sign of a rule that is not working.
Example. Cicero wrote to his brother in July 54 BC that the rate of interest at Rome had gone from 4 to 8 per cent because of the money being taken up for the consular elections. What does the remark establish?
That electoral bribery was large enough to move the price of credit in the largest financial market in the Mediterranean, and that it happened in a season, since Cicero reports a rate doubling in the weeks before a poll. It is also the right kind of evidence: an aside in a private letter to his brother, written for information rather than persuasion, with no reason to exaggerate and no audience to impress. A speech complaining about bribery would be worth much less, since complaint about bribery was a standard rhetorical move. Note what the figure does not tell us: how many voters were bought, or at what price, or whether the buyers got value.
Now you. The handbook advises Cicero to promise everything to everybody, on the grounds that a refusal is a certain injury and a broken promise a possible one. What does that advice tell you about the electorate?
Answer
That the voters who mattered were individuals the candidate met face to face and expected to meet again, not an anonymous public. The calculation only works if the injury of a refusal is personal and immediate and the discovery of a broken promise is diffuse and late, which is the arithmetic of a network of clients and go-betweens rather than of a mass electorate. It also suggests what a Roman voter thought he was buying: not a policy, since the handbook barely mentions any, but a claim on a man who might soon be in a position to help him.
What the voters could and could not decide
It would be wrong to end with the machinery alone, because the vote was not a formality and Roman politicians did not treat it as one.
Candidates lost. Assemblies threw out bills the Senate wanted and passed bills it hated, including the great commands of the 60s BC. A tribune with a popular measure could and did go over the Senate's head, and the fact that the Senate's usual response was to find another tribune to veto rather than to ignore the vote tells you the vote counted.
What the citizens could not do was set the menu. The choice at a consular election was among men of the same class, nearly all of them from families already in the Senate, presented by a presiding magistrate who could refuse a candidature. The bills came from magistrates. Debate happened at informal meetings, the contiones, where the presiding magistrate chose who spoke.
So the honest summary is that Rome had real elections without a real alternative, and the effect on the aristocracy was to make it competitive rather than to make it accountable. Since the reliable way to be remembered by voters was a war won, a triumph paraded, and games paid for out of the proceeds, the electoral system was a pump pushing ambitious men outward, towards commands, provinces and conquests.
Example. The four urban tribes held 4 votes out of 35 in the tribal assembly. Compute the share and say what it implies for a politician appealing to the city's poor.
Four of thirty-five is 11.4 per cent, against the 18 tribes needed to carry a vote. So a politician with the undivided support of the urban plebs, and of the freedmen registered with them, starts with roughly an eighth of what he needs and cannot get to a majority from that base alone. He must therefore either win rural tribes, which means the landowners whose clients fill them, or bring rural voters into the city, which means paying for their journey, or wait for a moment when the countryside is already in Rome. That is exactly the pattern of the popular politics of the late Republic, which peaks at festivals, in years of crisis, and around men rich enough to move people.
Now you. In 133 BC Tiberius Gracchus put a land bill to the tribal assembly, and the tradition says supporters came into Rome from the countryside to vote for it. Why is that detail worth noticing?
Answer
Because it is the mechanism by which a measure benefiting rural citizens could pass an assembly held in a single city. Rural tribes were normally decided by a handful of men present; when the beneficiaries themselves walked to Rome, the same 31 tribes that usually delivered the landowners' preference delivered the opposite. It shows the system was not rigged in the sense of being fixed, and it shows what changing the outcome cost: an organised movement of people over days of travel, which is why such moments are rare, memorable, and frightening to the men who normally controlled the count.
Where this leaves the Republic
Put the last three lessons together. Rome had a constitution that made action difficult, a career ladder with two winners a year, and an electorate that rewarded military distinction and was reached through networks of obligation. Every incentive in that system pointed the ambitious outward, towards a command against somebody.
Conveniently, Italy was full of somebodies. The next lesson follows the ambition out of the city and asks how Rome, having beaten its neighbours, did something no other ancient city did with the defeated: it turned them into more Romans, and thereby built the manpower that would fight Carthage.