Did Luke Fabricate a Census to Get Jesus to Bethlehem?

‘Tis the season for the same old skeptics trotting out those tired arguments about the birth narratives. Here’s biblical scholar and popular TikToker Dan McClellan rehashing the whole “Luke invented a census” spiel.

The story of Jesus’s birth, as found in Luke 1 and 2, is not historical….According to Luke 1, Jesus would have had to have been born, if not during the life of Herod, within about a year of his death at the latest. So, Jesus would have had to have been born by 3 BCE if Luke chapter 1 is historical. But chapter 2 introduces two critical problems.

The first is that the author needs to get Joseph and Mary from Nazareth down to Bethlehem so Jesus could be born there in fulfillment of prophecy. And the method they choose to accomplish that is to say they needed to go back to their ancestral hometown in order to register for this tax.

Seriously? Luke creating a huge empire-wide census to fulfill a prophecy about Jesus’ birthplace is beyond absurd. He doesn’t even mention the prophecy in Micah that the Messiah would come from Bethlehem. (Micah 5:2) Cooking up a census to shuttle Mary and Joseph to Bethlehem is a ridiculously complex plot device. If Luke wanted Jesus in Bethlehem against the facts, there were simpler ways to do it. Luke could’ve just started Mary and Joseph in Bethlehem, then moving to Nazareth. Or he could’ve made up a story of Joseph getting a job with his family in Bethlehem. Inventing an empire-wide census is like using a steamroller for a peanut.

Did Luke not know how censuses worked?

Dan continues:

Here’s the first big problem: Rome conducted a lot of censuses. They absolutely never required people to travel to their ancestral hometown to register. They registered in the towns where they lived and worked. Not only because it would not be feasible for the majority of people to travel to their ancestral hometowns, but it would entirely nuke the economy to have half of the Roman Empire just up and leave for who knows how many weeks. That was not how they conducted censuses. That absolutely did not happen in this case.

In one sense, Dan is right. Censuses weren’t this colossal empire-wide blitz squeezed into a few weeks. These censuses unraveled in these prolonged phases, zeroing in on specific chunks or regions. But living in this world, wouldn’t Luke have been in the know? He probably would’ve been enrolled in one at some point, or knew someone who had.

Furthermore, Dan is really reading things into the text here. Read Luke 2:1-4 for yourself. Luke isn’t portraying Augustus as saying, “Hey, drop everything and sprint back to your ancestral hometowns, pronto!” Dan is also assuming that Luke means everyone had to dive into their ancient family tree and rush to register at some ancestral hometown. But is that what Luke was getting at? In response to this style of objection, here’s what NT scholar Craig Blomberg has to say:

Luke 2:3 says that all went to be enrolled, each to his own city. One can understand from verse 4 why readers outside the Roman Empire unfamiliar with its practices could imagine his claiming that every Jew anywhere in the empire who could trace his lineage back to David had to go to Bethlehem, but Luke does not actually say that. “One’s own town” is elsewhere language for the city of one’s birth or one’s present or past residence (Josh 20:6; 1 Sam 8:22; 28:3; 2 Sam 19:37; Ezra 2:1; Neh 7:6; Matt 9:1; 13:57; Mark 6:4). Only a small minority of Jews in the first century lived somewhere other than the city in which they were born, so only they would need to travel for the census. Most tried to return to Israel for at least some of the annual festivals, so that would be a natural time for them to be “enrolled” as well.

The Historical Reliability of the New Testament, Craig Blomberg, pp. 132-133

Okay, next problem. Dan and other biblical scholars love to throw in our faces that ancient censuses never required people to return to their hometowns, but that’s not entirely true. There’s evidence suggesting that Roman authorities did enforce such obligations occasionally. Take Tacitus, for instance. In his Annals, Book VI, 41.1, he talks about the Clitae tribe in the Roman client kingdom of Cilicia. They were made to conform to Roman customs by declaring their property and meeting tax obligations, indicating situations where people had to return to their places of origin for census and tax purposes.

And there’s more. Archaeological digs uncovered a Greek papyrus from Prefect Gaius Vibius Maximus instructing everyone under his rule in Egypt to head back to their hometowns as part of a census. This discovery strongly supports the idea that the Roman administration did, at times, mandate hometown returns for census and taxation.

The old Quirinius conundrum…again

But Dan has more for us:

The other problem is even bigger, however. The author says that this happened when Quirinius was governor of Syria. And Quirinius became governor of Syria in 6 CE. But the census was not requested until 9 CE. And there’s a very good reason that that census was requested. And that was because the ethnarch of Judea, Idumea, and Galilee who took over after Herod the Great died was deposed in 9 CE. He was so incompetent Rome said, “You are banished, and we’re going to take over direct administration of Idumea, Judea, and Galilee.” So, even if a census had been conducted in Galilee for people living in Nazareth between 4 and 3 BCE, it would have been Herod Archelaus who was responsible for administering it. For people who lived in Nazareth, Quirinius, who was in Anatolia at the time anyway, who had nothing to do with the administration of Syria, wouldn’t even have been responsible for it.

So, the statement in Luke 2 that this happened when Quirinius was governor of Syria introduces an entirely irreconcilable historical datum. If chapter 1 is historical, chapter 2 cannot be historical. If chapter 2 is historical, chapter 1 cannot be historical. The two chapters are irreconcilable, and the story of Jesus’s birth is not historical in the Gospel of Luke.

So, Dan put it out there in a mea culpa moment—Archelaus actually got banished in 6 AD, not 9.

So that wipes out the idea of Archelaus having to be responsible for the census. Let’s pause and consider this for a moment. Apologies if what follows sounds a bit dry, messy, or dull, but well, ancient censuses are dry, messy and dull.

As I said earlier, Dan and I agree that censuses unfolded gradually. It’s possible that a count began in Judea before Herod’s death, and Quirinius later used it for tax collection in AD 6, perhaps assisting after Archelaus’ issues.

Historical records in The Deeds of the Divine Augustus (8.2–4) mention a census ordered by Augustus in 8 BC, involving about 4 million Roman citizens—massive for a population of roughly a million. With such figures, it had to be empire-wide. Considering slower communication and travel, regions like Israel may have taken years to complete the census. This suggests Jesus was born around 6–4 BC. If Matthew and Luke were making stuff up, it’s unlikely these details would align so perfectly.

Okay, now here’s where things gets a bit dicey. Luke 2:2 talks about a census during Quirinius’ time in charge of Syria. Matthew is pretty clear that Jesus was born during Herod’s reign, who died around 4 BC. The problem is, Quirinius didn’t step into that role until AD 6, a good decade after Herod the Great kicked the bucket. That leaves us with a pretty big discrepancy.

But here’s the thing: Luke knows that Jesus was definitely born during Herod the Great’s reign (Luke 1:5). He’s also aware of the taxation by Quirinius in AD 6 (Acts 5:37) as I mentioned already. Additionally, Luke provides clear indicators for the beginning of John the Baptist’s ministry and consequently Jesus’ ministry (Luke 3:1, 3:23). Following these cues, if Luke meant Jesus was born in 6 AD, he’d have been way too young. Luke says Jesus was around 30 when he started his ministry.

Okay, moving into the really nitty-gritty. The skeptical view of this passage overlooks the significance of that sticky little word “first” in Luke 2:2. If Luke merely meant to mention the census under Quirinius, why use the word “first”? Why not simply say, “This was the census during Quirinius’ time as governor of Syria” and stop there? The skeptics suggest “first” means “the first census in Judea,” but that’s not the only reasonable interpretation, especially if we take “first” as an ordinary adjective describing “census.”

Additionally, for us to trust Luke’s accuracy about the census, we don’t need one specific interpretation to be highly likely. Not to get too nerdy here, but what matters is for the disjunction to be probable—A, B, or C. And given that each of these ideas does have some real plausibility and aren’t just barely logically possible, the probability of that disjunction is reasonably high.

The scholarly debate around Luke 2:2 is vast and complex, but let’s stick to a few options to keep things from getting too tedious. One explanation comes from John Thorley, the other from NT Wright, and then another view comes from some older biblical scholars. Let’s start with Thorley’s view. In his paper The Nativity Census: What Does Luke Actually Say, he writes:

What Luke is actually saying is that this was the first census to take place while Quirinius happened to be governor of Syria, thus distinguishing it from the second census made when Quirinius was similarly governor of Syria in A.D. 6. Luke’s Greek surely allows no other sensible and unambiguous interpretation.

It might be objected that if Luke had meant the first of two censuses he would have written πρότερα. But in Acts 1:1 Luke uses πρώτον in exactly this way, and indeed πότερος was never obligatory in this sense. It must be said that several commentators have acknowledged this interpretation as a possible meaning of the Greek text, and the translators of the New English Bible inserted it as a footnote….

Let us then attempt a more precise translation of these two verses along the lines indicated above. Perhaps: “And about that time it happened that a decree went out from the Emperor Augustus that the census should be progressively extended to all parts of the Roman world. This was the first census to take place when Quirinius was governor of Syria.

Thorley points out that Luke’s accuracy about the decree or census in Judea around 8-5 BC and Quirinius’s rule in Syria remains a bit fuzzy. Although Augustus did carry out censuses, confirming Luke’s details has its share of challenges. Of course, there are lots of historical gaps in our knowledge. I’ll add that ‘ἡγεμών’ might not specifically mean ‘governor,’ suggesting Quirinius could’ve overseen a census in Syria without holding the governor title, as Josephus defines it. Maybe Quirinius held some kind of official role twice, or possibly Luke is mixing Quirinius up with Quinctilius Varus, who was the governor of Syria from about 6-4 BC. That might be an error on Luke’s part, but a small and somewhat understandable one. This challenges the idea of inerrancy, but it doesn’t mean Luke was just making stuff up.

Let’s check out option B. In his book, Who Was Jesus, the right Reverend Wright writes:

“It depends on the meaning of the word protos, which usually means ‘first’. Thus most translations of Luke 2.2 read ‘this was the first [protos] census, when Quirinius was governor of Syria’, or something like that. But in the Greek of the time, as the standard major Greek lexicons point out, the word protos came sometimes to be used to mean ‘before’, when followed (as this is) by the genitive case. A good example is in John 1.15, where John the Baptist says of Jesus ‘he was before me’, with the Greek being again protos followed by the genitive of ‘me’. I suggest, therefore, that actually the most natural reading of the verse is: ‘This census took place before the time when Quirinius was governor of Syria.’ This solves an otherwise odd problem: why should Luke say that Quirinius’ census was the first? Which later ones was he thinking of? This reading, of course, does not resolve all the difficulties. We don’t know, from other sources, of a census earlier than Quirinius’. But there are a great many things that we don’t know in ancient history.”

pg 127

If Wright’s right on the money, then Luke’s quick mention of the ἀπογραφή during Judas the Galilean’s time (Acts 5:37) doesn’t need any elaborate explanation. Luke’s short nod to the registration lines up with Josephus’s hint in Antiquities 17.2.4 about a loyalty oath to Caesar in Judea towards the end of Herod the Great’s rule—likely linked to a registration. No need to push Quirinius’s governorship back to 6 BC. Poof! All those apparent timeline clashes vanish.

Alright, let’s take a sec to look at Option C: Older historians like George Rawlinson and Alfred Edersheim suggest that the enrollment mentioned happened when Quirinius governed Syria. They argue that the term ἀπογραφή can mean (1) a registration or (2) a taxation involving a registration. There’s a possibility that a census started before Herod’s death and Quirinius used it for taxation in AD 6 after sorting out the situation following Archelaus being deposed. In Luke 11:28, a prediction of a famine is tied to the same term (ἐγένετο) used when Agabus predicted it and Luke later confirmed it under Claudius. Likewise, Luke might be saying that the census fully took effect during Quirinius’s governorship in Syria, later on. This could explain why there was a revolt in AD 6 but not earlier, if this census led to the taxation that caused unrest. It’s speculative, but worth considering. This view also has the advantage of wiping away all those pesky timeline problems.

The bottom line is that Luke believed in what he wrote in Luke 2:1-2—a genuine census in Judea, regardless of who was really overseeing it. The Quirinius-“first census” puzzle is a genuine subject for scholarly debate. Dan’s stance seems too rigid and doesn’t consider the whole picture.