Anyone who has studied New Testament history long enough knows the controversial Luke 2:1 census. The relevant verses (Luke 2:1-5) say:
“In those days a decree went out from Caesar Augustus that all the world should be registered. This was the first registration when Quirinius was governor of Syria. And all went to be registered, each to his own town. And Joseph also went up from Galilee, from the town of Nazareth, to Judea, to the city of David, which is called Bethlehem, because he was of the house and lineage of David, to be registered with Mary, his betrothed, who was with child.”
The problem is that Quirinius did not become governor of Syria until 6 AD, far too late for the events described in the nativities of Luke and Matthew to be reconciled with the above. Moreover, there was no such universal census that would’ve counted Joseph, who was not a Roman citizen, and Joseph would not have had to go to Bethlehem just because he was a descendant of David, who was born there: this is impractical, does not reflect Roman custom, and all of the Roman empire would’ve had to relocate itself if so.
But this does not reflect Luke’s often confusing expressions and narrative style. The verse that mentions Quirinius could easily be interpreted as “this was the census before Quirinius was governor of Syria.” (Luke 2:2). The census that Quirinius took in 6 AD was famous because it produced a massive revolt in Judea, and was a dating and reference point for many historians such as Josephus. Therefore Luke would have wanted to differentiate this one from it so that his readers are not confused, achieving quite the opposite.
Joseph Fitzmyer objects to this by (rightfully) pointing out that the grammar does not permit this interpretation (Anchor Bible Commentary, The Gospel According to Luke I-IX, p.401). However, this grammatical rule wasn’t followed half the time, least of all by Luke, who has numerous omissions, vague dating, and a mish-mash of a chronology and descriptive style. This alleviates the need to debate whether Quirinius could’ve been a governor of Syria twice, especially since there is no time for him to have been one around the birth of Jesus (c.5 BC).
The Universal Census under Augustus
The universal census was held by Augustus to count all citizens of Rome in 28 BC, 8 BC, and 14 AD. Of these, only 8 BC could qualify to coincide with Jesus’ birth, taking a few years to fully implement in a backward province like Syria (Palestine/Israel was a part of it at the time). We know from Josephus that whenever the Romans taxed Israel, even just the Roman citizens, Herod took it upon himself to tax everyone, which he loved doing for whatever excuse (this happened, for example, when a Roman commander arrived once, for example). So Joseph could’ve easily been “counted”.
The Need for Joseph to Return to Bethlehem
The question of why he has to return to Bethlehem has an equally simple answer: he had some kind of property there. But for whatever reason, possibly family disputes, he wasn’t practically in possession of it so relocated, or preferred to work/live in Nazareth. Nazareth was not the backwater village modern writers make it out to be (its location is actually unknown because the modern Nazareth has no “brow of a hill” that Luke 4:29 says it was built upon). Yet it must’ve been a relatively big enough city since Josephus tells us that the land in Galilee was fertile and therefore no settlement had fewer than 15,000 people. There are in fact more archaeological sites than we have names for in Palestine, so we cannot say there is a mistake here. In any case, Joseph had property in Bethlehem, and this is supported by Matthew 2:11. Joseph and Mary can’t find a place for Jesus into the “stable” according to Luke (the first floor of houses were where the animals lived), because of the issue of not living there. Under such circumstances, we know that Roman censuses actually demanded the return of the person to this location (for example, see here).
Finally, Luke 2:4 seems to suggest that Joseph went to “his” town because he was of the lineage of David. However, here is an example, just like in Matthew 27’s midrash on the death of Judas (plus the general conflation of prophets – e.g. Mark 1:2), of Luke giving a theological reason guiding a historical one which had different origins with respect to Augustus. What I mean by this is that Joseph returned to Bethlehem because he was of the lineage of David not with respect to the census, but with respect to God – God caused the census to happen which technically required Joseph to go to his property in Bethlehem, but ultimately because he was of the lineage of David so that the Messiah would be born there (Micah 5:2).
There is at least one author (I can’t recall his name or the book) who considered the passing reference to Quirinius’ 6 AD census in Acts 5:37 (Luke and Acts were written by the same person) to show that the Evangelist knew of the famous census everyone accuses him of mistaking and didn’t confuse it with the one in Luke 2:1. “Judas of the time of the census,” is easily a stock expression he took over from his sources or traditions, and he is in no way obliged to know it was in 6 AD.
However, since he was a careful historian (calls Pilate a “prefect” and not “procurator,” unlike his contemporaries such as Tacitus, himself a good historian, etc), and no ancient historian, Roman or not (e.g. Josephus), got such a major chronological marker wrong, it’s presumable that Luke did in fact know that this census occurred about 10 years after Herod’s death. The argument that Luke thought Jesus was born around this time (and that Herod died around 6 AD), is untenable because Herod was notable enough for Luke to have known when he would have died, and more importantly, he dates Jesus’ 30th year roughly to the 15th year of Tiberius (28/29 AD – probably the origin of the erroneous by ~5 years calculations for our modern AD/CE system), which means he could’ve in no way thought Jesus was born in 5/6 AD. If we subtract 30 from 28/29 AD we get 3/2 BC (no year 0, thus 1 year prior to 1 AD is 1 BC), which is much closer to 5/4 BC than it is to 5/6 AD. If he knew of Augustus’ 8 BC decree, he knew the census was in 6 AD and not earlier, and wasn’t talking about it.