Service in the fourth-century army was no longer a viable career path for most members of the Roman elite; and for recruits from a humble background, military life was less lucrative and more dangerous than it had been in the high imperial era.

To begin with the elite: wealthy men had often served in the officer corps of the first- and second-century legions. The military tribunes who assisted the commanding legate were typically young men of at least equestrian rank, and the legate himself was usually an ambitious senator. From the mid-third century onward, however, senators were effectively excluded from military careers - a fact of late antique political life that some senators, at least, bitterly resented (e.g. Symmachus, Or. 1.23), but were powerless to change.

Late antique commoners were not, of course, excluded for service. But they had less to gain and more to fear from it than their ancestors. For the auxiliaries who made up roughly half of the second-century army, military service was a path to citizenship. In the wake of Caracalla's grant of citizenship to all free Romans in 212, however, this perk vanished.

In real terms, the pay of late antique soldiers seems to have been considerably less than that of their mid-imperial predecessors. Legionaries serving under Trajan earned 300 denarii a year (plus a large bonus upon discharge). Although substantial amounts were deducted from their pay for equipment and clothing, this was a respectable salary, occasionally supplemented by donatives (payments celebrating an emperor's accession or other special occasion). Legionary wages were substantially increased from the early third century onward, but apparently failed to keep pace with inflation - at least in some places and periods.

Service, moreover, had become more dangerous. The late empire's enemies - the Sassanids in the east, various Germanic tribal coalitions in the north - were aggressive, and frequently successful against fourth-century Roman armies. In combination with recurrent civil wars (which were often the bloodiest of all), the threat of deployment cannot have enhanced the appeal of military service.

The size of the late Roman army is disputed, but it seems to have been, at least on paper, significantly larger than its mid-imperial predecessors. The need for more troops, in combination with the reduced benefits of service, probably accounts for the various laws against draft dodgers.

Joining the Roman army in the reign of Trajan promised a stable career and a respectable chance of surviving to enjoy a prosperous retirement. To many potential recruits, service in the fourth-century army did not.

The army remained a very important part of Roman culture - the epitomy of "manly virtue," and a position of high status and prestige. Late Roman soldiers (as well as Principate, i.e. that being 27 BC-284 AD, soldiers) held the status of honorati and as a result were above citizens in stature, holding tax breaks, legal privileges, and other privileges that the common citizenry did not. Comitatenses had more privileges than the limitanei, and the auxilia palatina probably had even more. So the features of the Roman army discussed in this thread that are often considered "lesser" than those of the Principate army turn out to be quite the opposite, although it should be noted that the reason the status and pay of the late Roman army had a wider gap from the common citizenry than in the Principate, was in part because of the need to attract more recruits due to larger demands for recruitment and a lack of access to recruits rather than an "unwillingness to serve."

The pay of the late Roman army, the key factor in attracting recruits, was also about the same as the Principate. At 4.5 grams the solidus was worth slightly less than the Aureus of say, Nero (~7.25 grams), but gold coinage was a greater incentive than silver or bronze coinage, which allowed the government to attract more recruits with less pay. This strategy was also used in the 7th century after the initial Arab conquests. The Roman army was paid entirely in Gold coinage from 398 onward to attract recruits, while the 1st and 2nd century army was largely paid in bronze sestertii, not silver coinage. Also, it's worth noting that even if the pay was considerably less, it was still more than the average citizen made in a year.

Anyways, the aureus was struck at 1/45 of a Roman Pound and the denarius at 1/96 of a Roman pound under Nero, meaning we can assume there were about 48 denarii to each aureus, give or take. That makes the average 1st century soldier's pay worth, give or take, between 5 and 7 aurei depending on the emperor (for Nero it would have been about 5, with pay at about 225 denarii), before deductions. The solidus was struck at 1/72nd a Roman pound or about 0.625 aurei. The average 4th century limitaneus was paid an annona of 3 solidi (about 1.9 Aurei) with the rest of their pay commuted in payments in kind (something the Principate soldier also received). Comitatenses were, however, given a pay of four annonae in gold coinage which in the 4th century was 12 solidi per annum, and one annona was usually deducted for payments in kind, for an equivalent of about 5.6 Aurei. Under Theodosius, the value of the annona was increased to 4 solidi per annum, and again under Anastasius to 5 solidi per annum. So the army of the 5th century was being paid even more, and the 6th century being paid more, with the average 5th century soldier being paid the equivalent of 2.5 and 7.4 Aurei for limitanei and comitatenses respectively, and the average 6th century soldier 3.1 and 9.3 Aurei respectively. And these are after deductions for payments in kind. Nor do they count the vestis militaris - a local tax collected by the soldiers to fund their clothing supply (which was also in place during the Principate).

It's often asserted that the limitanei were lower-quality "farmer-soldiers." The practice of limitanei taking on second jobs... was a thing. They weren't supposed to, and it wasn't really because they needed extra money but because they were being paid above the wage of the average citizen (who usually made the theoretical equivalent of less than 2 solidi per annum) and could put that money back into the local economy via their own businesses. The bigger issue was that they were having families near their stations and were being paid enough that they could afford to purchase land, and then instead of performing soldiering duties were helping their families with the land. They were allowed to purchase land, they just weren't allowed to purchase land in the provinces they were stationed in. Again, it wasn't because they needed the money, but because they had it. These families helped support the soldier, along with the local population in general, which made the burden of funding and supplying them far easier. But they weren't supposed to work that land instead of being full-time soldiers, hence the law about it passed under Valentinian III. Justinian's later "settlement" of limitanei into "farmer-soldiers" is a different matter, but also usually a misunderstanding of the situation regarding a shortage of funding for the military, and we'll assume for now we're talking only about the Roman army up through the fall of the west rather than get into the transition to the "early Byzantine" army.

Late Roman soldiers also served longer (20 years with two periods of re-enlistment to total up to 25 or 30 years), while Principate soldiers typically had varying lengths of service (16 to 22 years usually) before it was standardized to, more or less, the lengths of the late Roman system over the course of the late 1st through 2nd century. So they were paid more over their lifetime too.

Hugh Elton's 1992 study also estimates that the desire to volunteer for service did not entirely drop off, there just weren't enough to meet the full demand of military recruitment. There isn't a lot of recorded evidence for volunteers, but he notes that there are several 4th and 5th century examples who are stated to have volunteered for service, including the emperor Majorian. He also notes that the practice of mutilation to avoid service was still an extreme case, and was not really as much of an issue as people think. Deserters had always been a problem to over the Roman period and with any pre-modern army, and typically that issue varied in severity depending on the success of the army at that particular time.

The primary issue with recruitment was the commutation of recruitment by landowners to payments of cash to the emperor, called an adoratio. The adoratio became so common that the Romans eventually started using it to call for emergency taxes. The last recorded call for recruits that specifies persons for service was in 428 AD under Valentinian III, while the last call for recruits used to levy an adoratio was in 444. The other major economic issue was the fact that the average citizen had a value of less than 2 solidi per annum. Families had hereditary service requirements, and in 386 Valentinian II passed a law that allowed for 6 families worth less than 2 solidi per annum to fund a single recruit between them, rather than each fulfill their hereditary service requirement individually. Families worth 3 solidi per annum or more still had to fund their recruit individually. This practice was reinstated under Charlemagne in 808, and Nikephoros I in 809, becoming the "Theme system."

(1/2)