After having spent some time studying tax farming in the Ottoman context, I'm now curious about its functioning in other societies as well. What is the state of our knowledge with regard to the Romans?
I'd be interested to hear in general how historians of Rome have approached it, but I do have some specific questions regarding the intersection of tax farming and politics. How were tax farmers assigned, and did control over tax farming contracts become a form of patronage for Roman political leaders? Were there any groups that attempted to create monopolies or cartels to control tax farms on a large scale? With regard to finance, did the Romans ever use tax farms as a way of anticipating revenue, by selling contracts for future collection in advance (and if so, what effect did this have on their distribution)?
General description
We know quite a lot about the Roman economy, especially in the Late Republic and the Empire, but unfortunately the tax-farming aspect is one of the areas where we know a lot less than we would like. There is no single quintessential source for the practice, and we have to piece it together from mentions across time and genre. Cicero talks about it a bit during his prosecution of Verres for provincial governing misconduct; Livy mentions it here and there; et cetera. What sources we do have on it are often distorted or flat out wrong, in addition.
In essence, the Roman state contracted out the burden of collecting revenues (of various sorts) from the provinces (and sometimes within Italy) to groups, what we might call corporations, called publicani. The publicani were, from the Punic Wars onwards, invariably equites, "Roman knights," the wealthy but non-senatorial class at Rome. Many of them were "Italians" rather than Romans before the Social War in the early 1st century BCE, though having a Roman citizen at the head of the groups might have been a requirement for Roman contracts before the 1st century BCE—on that I'm not sure. If so, I'm sure they found ways to cope with that quite easily. The Publicani had their fingers in many pots: tax-farming, yes, but also shipping, mining, provisioning of Roman armies, the slave trade, finance, banking, works projects, etc. All of these activities were, theoretically, barred from Senators and Roman magistrates. This ostensibly prevented direct exploitation of provincials by Roman officials, though of course there are always means to circumvent such rules. According to the conservative, ancient customs of the Romans, senators should derive their wealth from land holdings and should NOT be engaged in negotium, business. [edit: left out the very important NOT in that sentence: Senators were NOT to be engaged in business; it was beneath them].
Like modern corporations, individuals formed companies to pool resources, raise capital, and share risk. Each contributing member had a stake tied to contribution, like stock holdings (partes). The partes could be transferred and traded, and apparently even sold (see Cicero P. Vat. 12.29, partes illo tempore carissimae 'shares very valuable at that time'). The companies had centralized leadership back in Rome and field managers out in the provinces. Such companies (called a societas, or socii) are known from the Punic Wars at least, and are probably older. At least some of them seemed to have public-facing "offices," like the ones we find at Ostia in the famous Piazzale delle Corporazioni. And of course the publicani had small private armies of minions to carry out their tasks, including the actual gathering of taxes, surveying, sailing, selling, and what have you. These employees were called familia publicanorum, which I love. It has a distinctive mafia flavor to it. It's just business. The activities of the publicani and their minions are everywhere in the Late Republic and early Empire, often well beyond the established borders of Roman dominion. They formed a massive but often (to us) invisible substrate of Roman/Italian presence from Britain to Arabia, the Sahara to the Black Sea. For the structure and functioning of the societates, see for instance Cicero's offhand comments in pro lege Manila 2.6 and pro. Rab 2.4.
Why farm out the taxes?
Holland (in Money in the Late Roman Republic 2007) identifies three compelling reasons for utilizing the publicani to farm tax revenue. First, it shifted the burden of bureaucratic complexity onto the companies. The Roman state, especially provincial administration, is well-known for its peculiarly-narrow bureaucracy, and this is one of the major reasons. Second, it provided a reliable and predictable revenue stream for the state, since the publicani "bid" on the sum to be collected and provides sureties up front. They shouldered the risk of the actual collection and they were responsible for dealing with any unforeseen variables. Third, by using the publicani as middlemen, the state received only cash, never kind. If a provincial region was paying their requisition in turnip harvest, it was the responsibility of the publicani to convert the turnips into cold hard cash for the state (or, in some cases, to recoup the cash they had already fronted the state). How or why regions paid in cash or in kind is another topic, and it seems to have varied greatly. We have evidence that in some places, farmers got a premium on their goods in kind if those goods were earmarked for markets in the city of Roman, which is fascinating as far as privileged economic zones but not relevant here.
How were they assigned? And was it political?
The censors (overseers of the senate rolls) were responsible for managing the distribution of contracts via bids submitted by the manceps of each societas (a sort of CEO), usually on 5-year terms (see Cicero de Leg. Agr. II.21 and Verr. III.7; Livy 24.18.10). This was at a public auction in Rome. The censors set the terms and made the final decisions, again in public, though we hear that the arrangements could be initiated or altered by the People (via tribunes) or by the Senate. Interference seems to have been rare, from what I can gather. That the contracts were let out by the censors suggests that, per your question, the awarding of contracts was not especially married to political favor. The censors were the highest ranking senators of their time, ex-magistrates, usually old farts, and typically long past their campaign days. They were supposed to be (and probably though of themselves) as being above the squabbles of politics.
We gather that the contract winner either deposited the agreed-upon sum (in bullion) into the Roman treasury, or else provided adequate surety for the amount. Either way, the State got theirs up front, and what was exacted from the provinces was in the hands of the agents. The art was bidding high, but not impossibly high. The exact opposite was the case for bidding on the public contracts, such as the contract for a new road. In that case, bid as low as possible and hope it can be done for even cheaper. See, e.g. Livy 23.49, or 24.18 for a case of taking contracts without immediate payment in times of war.
Under the Empire, there is a gradual shift away from the engagement of publicani for public projects. Most associate this with the Emperor's desire to have more direct control over the operations via appointed officials (procurators, prefects, Imperial freedmen, etc). By Late Antiquity, we find that many areas once dominated by publicani are now wholly in the hands of Imperial corps of bureaucrats and very much dependent on the Roman Imperial governmental system to function. When that machinery ceases to exist, those industries ceased to function immediately (for instance, the construction and maintenance of aqueducts, roads, harbors, canals), and there was no pathway for the mountains of institutional knowledge to flow to new recipients.
As far as politics: the publicani themselves were of course never above playing politics. They were individuals and groups but often worked in concert for their common goals. Cicero makes few distinctions between the equites as a class and the publicani. There are numerous examples of the "business class" exerting their political will overtly or behind the scenes in Rome, variously in tandem with or at odds with the senatorial ruling class. In 67, the equites (read: publicani) were instrumental in getting Pompey installed into a special military command against the pirate menace, with near-unprecedented powers. The pirates, after all, were getting out of hand, and business all over the Mediterranean was suffering.
Monopolies or cartels?
I do not know the answer to this question. It would make perfect sense that ultra-rich, politically-powerful individuals (like Crassus, or Atticus) would be able to manipulate the playing field of the publicani. I also do not know enough to talk intelligently about how the Roman system of patronage intersected with the publicani in any period. There is no obvious mechanism for fair play except competition itself inherent in the bidding process. I do not know about this topic to say more, and maybe the evidence is lacking, but in my opinion, near-monopolies or consortiums were totally possible and even likely.
Publicani as emergency revenue
I don't know if the state engaged in futures of public contracts. I tend to doubt it. But we do have good literary evidence that the publicani "came to the rescue" of the Roman State on at least one occasion, and probably more. During the Second Punic war, when, after Cannae, the Roman coffers were nearly empty, Livy (24.18) tells us that publicani offered the state a substantial loan. This is fascinating, in that it indicates a robust system of publicani or their functional predecessors in the 2nd century, and also because it shows just how much capital such groups were able to raise seemingly at will, even in times of great deprivation.