The census in Luke 2 is a gigantic topic on which much ink has been spilled. I certainly had to deal with it in my series on the Virgin Birth, but I'm trying not to write a treatise! This post contains some extra notes on the subject that I didn't include in my recent video, in the interests of keeping the video streamlined and digestible. Here also is my recent debate with atheist Jonathan Pearce on the Unbelievable show.
In my Youtube video about the census I make the following points:
1) Luke is an historical source in himself, at least as credible (based on track record) as Josephus. The fact that we have no other source for a census in Judea at this time is thus the merest argument from silence, and an especially poor one. Luke is giving us information. There is nothing about Luke's being a Christian author that makes him likely to be unreliable about a boring historical matter like a census. This point goes beyond saying that we should "give Luke the benefit of the doubt" based on his track record elsewhere, though that's true as well. The mere absence of a census in this time and place in the relatively meager set of non-biblical historical literature that we have for that time period does not constitute a strike of any significance against Luke. We can learn about this census from Luke.
2) It's outright false that Luke and Matthew contradict each other about when Jesus' birth took place, with only Matthew placing his birth at the time of Herod the Great. Luke says the same thing. (Luke 1:5)
3) Skeptics will say that Luke's census is improbable to the point of being historically impossible, even on its own terms. They are wrong about that, and they get there by insisting on an overly wooden reading of Luke and an exaggerated idea of what the census would have involved.
4) Skeptics will say that there is only one possible meaning of Luke's reference to Quirinius in connection with the census and that all other suggested translations are attempts by Christian apologists to wriggle out of admitting that Luke was wrong. They're wrong about that, too. What Luke says about Quirinius and "the first census" is genuinely difficult to translate and interpret, which is why there is legitimate scholarly debate about it.
Here I want to add a couple of points to #3 and #4.
Concerning #3, one claim that you will here is that Rome would never order a census (either for purposes of counting or for purposes of taxation) under a client king such as Herod the Great. This is a really strong claim, and there is little to back it up. It's mostly just an assertion, based on the fact that client kingdoms did have some measure of independence. But it's not as though we have a contemporary statement anywhere that the Romans would never meddle in taxation in a client kingdom or would never order a tally of the people in a client kingdom.
On the contrary, here is some evidence that Rome would sometimes do so: In the 30s A.D., as Tacitus tells us (Annals, Book VI, 41.1), a rather war-like tribe (the Clitae) residing in the Roman client kingdom of Cilicia was "pressed to conform with Roman usage by making a return of their property and submitting to a tribute." They were originally from the mountainous region of Cappadocia, and they retreated there and fought. The client king, named Archelaus, required the help of the Roman legions to defeat them.
As it turns out, there were several rulers about in the 1st century B.C. and 1st century A.D. named "Archelaus." This one (whom Tacitus calls "Archelaus of Cappadocia") was not the same person as the Archelaus, the son of Herod the Great, whom I've mentioned elsewhere (see Matt. 2:22). Making things more confusing, the Archelaus ruling the client kingdom of Cilicia, mentioned by Tacitus, was also not the "Archelaus of Cappadocia" who had died about 17 years previously. That was his father, who actually did rule Cappadocia. (Are you confused yet?) I bring all this up because the atheist blogger Jonathan Pearce (who debated me on the Nativity) has claimed that the Biblical Archaeology Report has blundered horribly by mentioning this tribute/census of the Clitae as evidence that you could have Roman censuses in client kingdoms. Pearce assumes that they are referring to a census made in Cappadocia after it was no longer a client kingdom, when Archelaus of Cappadocia had already been dead for years. But he's mistaken. The requirement to "conform to Roman usage" was indeed made within a client kingdom, in the year A.D. 36, when that client kingdom was ruled over by a different "Archelaus of Cappadocia," the son of the one Pearce is thinking of, and (to make matters more confusing) the client kingdom in question was actually Cilicia rather than Cappadocia! The confusion over this obscure fact partly arose because the Biblical Archaeology Report cited a secondary source rather than citing Tacitus directly. I was lucky enough to find the Tacitus reference in this really fascinating article about the census by John Thorley (hat tip to Jason Engwer for recommending it), and I chased it down from there. (Note: You can get a free on-line account with JSTOR for independent scholars that lets you read up to 100 articles per month.)
This is a cautionary tale in a lot of ways: It illustrates the complexity of historical reality (always something skeptics and some Christian biblical scholars need to be reminded of). It illustrates the plausibility of theories that there were multiple people by the same name. (This comes up in discussing other supposed Gospel "errors.") In this case, Wikipedia even calls the Clitae a "Cappadocian tribe," which is confusing and interesting. How could there be a Cappadocian tribe in Cilicia? Well, you know, history is complicated! Anyway, all of this also illustrates the value of tracing things back to original sources.
And bringing us back to the argumentative point: We absolutely should not be doing a priori history about what "wouldn't ever happen" in a client kingdom. We should discover what it meant to be a "client kingdom" in regards to tribute, census, taxation, etc., by reading historical sources (including Luke). The phrase "client kingdom" isn't some kind of talisman that automatically entails the conclusion the skeptic is going for. That's not how history is done. So...yes, Rome under Augustus could certainly have ordered that a client kingdom (or maybe even several of them) must count their people, or their property-owning people. Herod would have had to agree to carry this out or to allow a designated Roman authority to carry it out. As Biblical Archaeology Report notes, Augustus ordered after Herod's death that Samaria didn't have to pay as much tax to Archelaus, Herod's son, because they hadn't joined in a revolt. Even though Archelaus was confirmed as a ruler under Rome and was supposed to have the tribute from the Samaritans himself, Caesar altered the amount. This is indirectly relevant in that it shows how Rome tweaked taxation under client rulers.
In fact, Thorley suggests that that is what Luke is saying when he says that "in those days there went out a decree from Caesar Augustus that all the world should be taxed [registered]." Thorley's idea is that this means that at that general time period Augustus decided to tell the various provinces on the edge of the Roman empire to register/count their inhabitants, this extension of census-taking to all parts of the Roman world being a new practice. Of course, there could be plenty of reasons for this. It would be useful for taxation, for one thing, even if some part of the tax went to the local client king. (Miller, link below, mentions the suggestion that Rome might have wanted to assess the region when a client king was getting older.) The story of the Clitae in Cilicia shows that this is not merely theoretical. A tribe in a client kingdom was required to make a tribute of a portion of their property according to Roman usage, which would have required that the property be counted, and the client ruler had to attempt to carry out this order. While Tacitus doesn't say whose idea this census was (it could in theory have been the idea of Archelaus II himself), he certainly doesn't say that it was not Rome's own idea, and the fact that the legions helped to enforce it certainly shows their involvement. For more on this tribe see Glenn Miller's extensive discussion here. I haven't even had time to read it all and therefore am not endorsing everything he says, but it contains a lot of information. (HT to Jason Engwer for the link.)
It's also worth pointing out that Herod's relationship with Augustus as a client king was not always strewn with hearts and flowers. Some time between 12 and 9 B.C., Herod fell into significant disgrace with Augustus over his treatment of the Nabateans. While he was supposedly reconciled to Augustus, such a reconciliation wouldn't have meant that they both forgot the recent unpleasantness, even if they were formally friends again by the time of Jesus' birth, Augustus had made it quite clear recently that he considered himself fully empowered to interfere in Herod's management of his affairs.
Rome had a passion for counting people, not to mention taxing them. Augustus proudly talks about several lustrum censuses he did, including one beginning in 8 B.C., and how many Roman citizens he counted in those censuses. I should clarify here something that I was not clear enough about in my debate with Pearce: In itself, a lustrum census was for purposes of counting Roman citizens, not all inhabitants per se. There doubtless were Roman citizens in Judea, but if the census at the time of Jesus' birth was related to the 8 B.C. lustrum as it came around to Syria, that would be an extension of its independently known purpose, since Joseph was probably not a Roman citizen. But as Thorley points out, that isn't in itself implausible.
Concerning #4, here are a couple more points (which may or may not have been covered in the video):
Skeptics insist that Luke must be saying that this census took place while Quirinius was governor of Syria and that, moreover, he is saying that it was the 6 A. D. census. So there you go! Luke was wrong! Because Jesus wasn't born in 6 A.D.!
But that claim has all sorts of problems of its own. It runs afoul of Luke's clear assertion that John the Baptist was conceived during the time of Herod the Great (Luke 1:5). It runs afoul of Luke's clear knowledge of the 6 A.D. census at its own time period (Acts 5:37). It runs afoul of the fact that Luke has all kinds of definite indicators of the time of the beginning of John the Baptist's ministry and hence Jesus' ministry (Luke 3:1, 3:23), and that on these indicators Jesus would be much too young if Luke thought he was born in 6 A.D. In other words, it just doesn't make sense within Luke's own corpus as an interpretation of what Luke is saying about Quirinius and the census in 2:2. Whatever else is going on, that isn't what's going on. Even if one thinks Luke is incorrect here in some way, that isn't what he's saying just as a matter of interpretation of Luke.
Moreover, the skeptical interpretation leaves out the meaning of the word "first" in Luke 2:2. If all that Luke meant to say was that this was the census taken under Quirinius, the one and only, why didn't he just say, "This was the census made when Quirinius was governor of Syria," full stop? Why include that pesky word "first" (or whatever it should be translated as)? The skeptics apparently think it means "the first census in Judea," but that is far from being the only reasonable interpretation, even if we take "first" to be just an ordinary adjective modifying "census."
Let me also add: In order for us to be justified in thinking that Luke is accurate about the census (in particular), it isn't necessary for any one specific possibility to be probable. What is needed is for the disjunction to be probable--A or B or C. And given that each of these ideas does have some real plausibility (it isn't just barely logically possible), the probability of that disjunction is reasonably high. If one allows for Luke to be right that Jesus was born during a census or registration but wrong about the specific Roman governor/hegemon, through confusing two Roman names (see below), a minor error, then the probability of the disjunction is higher still. So when I mention in the video that this or that is possible, I am not saying either a) that being possible means being probable or b) that these are individually just barely possible. In fact, common sense shows us that many of these ideas that I've suggested about the nativity (such as Joseph's having a connection to Bethlehem stronger than just being descended from David) are entirely plausible, and they are the kind of thing that we invoke in ordinary life all the time to explain what someone says when we have incomplete information. Jonathan Pearce, my atheist opponent in my recent debate, repeatedly states that "apologists" invoke the idea that to be possible is to be probable. This misunderstands the entire point. Moreover, it's particularly ironic that he should repeat this criticism so often, since he himself invokes, and treats as highly probable, extremely implausible theories, such as the idea that Luke 1-2 were added to the Gospel later on or the idea that Luke is secretly trying to make a reference to Psalm 87:6 by making up/moving the census.
Thorley's suggestion, which I'm inclined to endorse as my "first line" of translation, is that Luke is saying that this was the first census (of two) made when Quirinius was "hegemon" of Syria. Again, that "first" has to mean something. We can't just leave it untranslated. Luke is trying to communicate something. I add, which Thorley doesn't talk about, that "hegemon" doesn't have to mean "governor" in the technical sense and that Quirinius could have been in charge of a census in Syria without being governor in the sense that Josephus talks about when he lists the governors. But Thorley, (refreshingly) taking Luke to be an historical source, also says that for all we know Luke is more accurate than Josephus here and knew of a short, earlier governorship of Quirinius wedged between those listed by Josephus, which is also possible.
There is some controversy over the suggested translation, "This census was made before Quirinius was governor of Syria," though it would certainly be a simple way to fit all the data together (always a good thing in an historical hypothesis). I find its simplicity attractive. That translation, by the way, would mean that Luke is particularly accurate here. Here is Daniel Wallace making a case against it. But N. T. Wright endorses it (Who Was Jesus? p. 89). (So did a whole roster of other scholars who were no slouches in Greek, including T.R. Birks. I owe this reference to Tim McGrew.) Maybe we should let Wallace and Wright duke it out on this one. In any event, the contemptuous skeptical dismissal of the "before" translation as a desperate apologetic expedient is unwarranted. It deserves consideration.
Another worthy contender is, "This enrollment was first completed (i.e., used) when Quirinius was governor of Syria." This is Calvin’s suggestion, endorsed by Beard, Rawlinson, Edersheim, and numerous other scholars. As Paul Maier points out, it took forty years to complete a census in Gaul around this time, so it could well be that a count was made or begun in Judea before the death of Herod and that Quirinius only made use of it to collect tax in A.D. 6 when he came to clean up the mess after the death of Herod the Great's son, the tetrarch Archelaus. Luke 11:28, referring to a famine that at that time was merely predicted, uses the same Greek term (egeneto). Agabus predicts the famine, and Luke comments that it happened, came to pass, etc., under Claudius, in a future time. Similarly, Luke could be saying that this census came fully to fruition when Quirinius was governor of Syria, later on. (I owe information about this option to Tim McGrew.) It occurs to me that this could explain why there was a revolt later in A.D. 6 but not at this time, if this was the count and that was the taxation based on the count. This is just a conjecture but is worth throwing into the mix.
Thorley suggests that Luke may have been mistaken, but only in a narrow sense. Luke may indeed have said, "This was the first [of two] censuses made when Quirinius was governor of Syria," thinking that Quirinius was hegemon of Syria twice, based upon a mistaken memory or reading of the name Quinctilius, since the hapless P. Quinctilius Varus was, according to Josephus, governor of Syria around this time. That would indeed be an error on Luke's part, but a very limited and to some extent understandable one, and it certainly wouldn't at all mean that Luke invented the census. Indeed, Luke's very attempt to nail down the relationship of this census to the one that he knows about later under Quirinius shows an extremely Lukan concern for literal history.
In fact, the whole idea that Luke made up the census (or moved Jesus' birth to much later), as I point out in the video, is fairly absurd. It is using a steamroller to crack a nut. All that Luke had to do, if he wanted to "make" Jesus be born in Bethlehem contrary to fact, was to have Mary and Joseph start out in Bethlehem and later travel to Nazareth. There was no need for him to invent the idea that Mary was from Nazareth and that they had to travel from there, while she was pregnant, down to Bethlehem and then back to Nazareth. And to invent a Roman census for such a purpose would be a wildly exaggerated plot device. Luke's deliberately connecting it falsely with Quirinius and placing it at a date that is in great tension with all of Luke's own other date indicators is overwhelmingly implausible. Why do a thing like that? Luke didn't have to mention Quirinius at all if he was inventing a census out of thin air.
Luke shows not the slightest awareness of any Old Testament passage that is fulfilled by Jesus' birth in Bethlehem. He may or may not have known of Micah 5:2. I think it's a good principle not to attribute theological motives to the evangelists that they say nothing about. They generally aren't shy about mentioning OT parallels or fulfillments of prophecy, so why invent private intentions for which we have no textual evidence?
All of our evidence points to the conclusion that Luke sincerely believed what he said in Luke 2:1-2. And there are plenty of reasons to think that, as a reliable historian, Luke is telling us about a real census that really took place in Judea at the time, whether or not Quirinius was in charge of it. To say that we wouldn't think there had been such a census if it weren't for Luke is no real criticism. There are plenty of historical events that we wouldn't think happened if it weren't for the historical document (sometimes a single document) that mentions them! That's how history works.
Have a look at the video, and be sure to subscribe to my Youtube channel if you haven't already!