Tertullian states that censuses (plural) were conducted in Palestine around the time of Jesus’ birth. There are two viable candidates for the census described in Luke 2. One is the census of 8 BC (the second of three censuses ordered by Augustus during his 41-year reign), documented by an inscription found at the temple of Augustus in Ankara, Turkey.
The other option is a combination of census and oath of allegiance ordered to celebrate Augustus Caesar’s silver jubilee on February 5, 2 BC. This celebration marked the 25th anniversary of Augustus’ elevation to supreme power by the Senate and people of Rome. It was also the 750th anniversary of the founding of Rome. At this celebration, the Senate conferred upon him the title Pater Patriae (“Father of [his] Country”). The year before, Augustus sent out a decree requiring “the entire Roman people” throughout the empire to register their approval for the bestowal of this honor (T. Lewin, Fasti Sacri  135). This registration was required of all Roman citizens and others of distinguished rank among Rome’s client kingdoms such as Judea.
Josephus substantiates that such an oath of allegiance was required in Judea at this time. In Antiquites 17 he mentions that at this time “all the people of the Jews gave assurance of their good will to Caesar, and to the king’s government.” He further notes that there were six thousand Pharisees who refused to swear the oath. By this time, having already fallen out of favor with Rome, Herod was scrambling to stay alive politically. The refusal of a large contingent of Pharisees to pledge allegiance to Augustus would have been a major embarrassment for Herod. That may have been precisely the Pharisee’s calculation. It certainly would not have helped his mood should visitors from the east bring word of one “born king of the Jews” some time shortly thereafter.
A decree going out from Caesar Augustus in 3 BC fits nicely with the early patristic evidence. This date, plus Josephus’ mention that an enrollment was in fact being taken in Israel a year or more before Herod’s death, lends weight to the idea that the lunar eclipse before Herod’s death was the one that took place on Jan 9, 1 BC. If Herod died the following year after the Dec 29 eclipse, the census in Israel would have been far too late for Augustus’ jubilee. A late census is a possibility, bureaucracies being what they are, and Rome had every reason to be thorough about this in as volatile a region as Judea. The earlier date, however, yields a more plausible time line.
Obviously, the recording of oaths (where people ascribed their names) was a type of registration. That is what Luke said the census was: it was an enrollment of people. This enrollment would have included Joseph and Mary even though they were not Roman citizens. As “royal claimants” they would have both been especially singled out to pledge their loyalty to Augustus.
There are two problems, however. First, is it credible that the Romans would require people to be registered in their hometowns? Wouldn’t this create an even bigger bureaucratic nightmare? Perhaps it did, but there is archeological confirmation for the practice. A provincial census decree from Egypt dated AD 104 required absentees to return to their hometowns to be registered. The decree reads,
Gaius Vivius Maximus, Prefect of Egypt [says]: seeing that the time has come for the house to house census, it is necessary to compel all those who for any cause whatsoever are residing out of their provinces to return to their own homes, that they may both carry out the regular order of the census and may attend diligently to the cultivation of their allotments.
There is no reason the Romans would not have followed a similar procedure a century or so earlier in Judea.
The second problem relates to the governorship of Publius Sulpicius Quirinius in Syria. According to Luke 2:2, the “Christmas census” seems to have taken place while Quirinius was governor of Syria. It is admitted by everyone, however, that Quirinius only became governor of Syria in AD 6. In the time frame in question, Saturninus was governor (proconsul) of Syria—as Tertullian admits. Saturninus left this post early in 2 BC to be replaced by Quinctilius.
There is in fact evidence that Quirinius may have been active in the government of Syria much earlier, perhaps as a special representative of the emperor, a procurator or legate. Some scholars appeal to this possibility in order to “rehabilitate” the text of Luke. By this theory, Luke’s Greek wording is taken be more generic: “while Quirinius was ruling Syria” (Lk 2:1). Quirinius was made consul in 12 BC, one of the two most powerful men in the empire behind the emperor himself. In this capacity, he served as supreme military commander of the eastern provinces in the last decade before Christ. He especially distinguished himself during a five-year war against the Homonadensian tribe of Cilicia, roughly dateable to 10–5 BC. According to Justin Martyr, at the time of Jesus’ birth Quirinus was procurator of Judea (a district within the larger province of Syria at the time). It is thus not impossible that Quirinius was in Palestine conducting the census while Saturninus was the actual governor of Syria. In fact, Tertullian states that the “Christmas census” took place while Saturninus was governor of Syria, and places it in 3/2 BC!
There is, however, a simpler solution to this chronological puzzle (for which BK has recently provided a concise summary at the CADRE Comments blog). The Greek text of Luke 2:2 is as follows: haute apographe prote egeneto egemoneuontos tes Syrias Kyreniou. It is generally translated something like, “This first census came about while Quirinius was governing Syria.” It is possible, however, to translate prote not “first” but “before” or “prior.” The same word is translated in the NT as “former” (Ac 1:1). Nigel Turner, a Greek scholar of the first magnitude, suggests that a better translation of Luke 2:2 would be, “This census was before the census taken when Quirinius was governor” (Grammatical Insights into the New Testament, 23-24). If this reading is adopted, it makes any attempt to locate Quirinius in Syria at the birth of Jesus irrelevant. Luke’s point was to differentiate this prior census from one perhaps more familiar to his readers.
The census of 8 BC is a fact of history. As a subject of Rome, it may well have required Joseph’s participation as a male and thus presumably a property-owner. This enrollment would not, however, have demanded Mary’s presence, and it is unlikely that Mary would have traveled with him in her ninth month of pregnancy if she did not have to. A better option for the “Christmas census” would be the census and oath of allegiance ordered in preparation of Augustus’ silver jubilee in February of 2 BC. This census is also a fact of history. If my reading of Josephus is correct, this registration was taking place in Israel 12-15 months prior to the death of Herod, most likely some time between November 3 BC and February 2 BC.
Next: The Star of Bethlehem