There are many “gotcha” verses that Bible critics like to point out as evidence in their mind that the Bible can’t be trusted, but most of them do not present the historical quagmire that Luke 2:2 does. This verse begins the opening of Luke’s chapter describing the birth of Jesus, and it reads:
“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.”
Put simply, the problem with this is that the documented census during the governorship of Quirinius happened in 6 AD (the annexation of Judea, cited by Josephus in his Antiquities 18.1). To make matters worse, this is cross-referenced by Luke himself in Acts 5:37 with Judas the Galilean, a Jewish revolutionary who led an uprising to the Roman tax census in 6 AD and failed. Why this is a problem is that Luke describes the census in chapter 2 of his gospel as taking place when Mary was “with child” — as in pregnant.
I have documented why Jesus was born in 3 BC and why Herod actually died in 1 BC, not 4 BC (as critical scholarship suggests) — however this does not solve the problem of Luke 2:2. In fact neither date solves the problem, because you have anywhere from about 6-9 years of mismatch between the census in 6 AD and the “registration” Luke is describing. And if Luke is confusing names or places or dates, then that creates a problem to be reconciled.
So this is an important issue to resolve. However there are more difficulties as we examine the possible ways to reconcile this, and they create the need to be very conservative and very careful in our understanding, while maintaining the inerrancy of scripture. Examining each helps us paint the picture of what’s possible and what isn’t.
OPTION 1: “FIRST” ACTUALLY MEANS “BEFORE”
The first possibility is that what Luke actually meant in the original language is “before the registration” as in “This was before the registration when Quirinius was governor of Syria.” The argument places whatever census took place before Quirinius’ governorship, and scholars like NT Wright rely on John 1:15 as a parallel text to support this original language argument, where John the Baptist says, “‘He who comes after me ranks before me, because he was before me.” The word “before” there is the same word as in Luke 2:2 where it says “first” (protos/prōtē).
But the problem here is that Luke 2:2 is grammatically different. The original language text is reproduced below from the interlinear:
αὕτη ἀπογραφὴ πρώτη ἐγένετο ἡγεμονεύοντος τῆς Συρίας Κυρηνίου
Here the word prōtē (πρώτη) is a feminine nominative (an adjective that labels the subject), and it agrees with “registration” (ἀπογραφή) as a predicate adjective — “this registration was first.” In Luke’s sentence, the word “first” directly describes the “registration” itself — essentially saying, “this registration was the first one.” What follows is a self-contained phrase that sets the time for this event to have happened: “while Quirinius was governing Syria.” So because these words agree, the Quirinius qualification demands that prōtē means “first” not “before.”
So because the sentence structure is completely different from the verse in John, you can’t just copy the meaning of “before” from one to the other. That’s the issue. The context and grammar make the word do different things in each situation. The technical explanation is because in John the genitive (a form that shows possession or a close connection between two words, like “me” or “mine”) is the complement of protos (πρῶτος), and in Luke it belongs to a separate participial clause. A simpler way to put it is that the word “before” works in John’s gospel because it directly connects to a noun (“he was before me”). Here, “me” is the genitive pronoun and it’s being complemented by protos directly (“before”). However, in Luke’s gospel, that same word complements the “registration”, which itself is qualified by a completely separate action phrase (Quirinius governing Syria). So “before” doesn’t work and it has to mean “first” instead. That means John 1:15 cannot be used as a grammatical parallel for Luke 2:2.
A renowned New Testament scholar named Joseph Fitzmyer, who has done a lot of work on Aramaic, Dead Sea Scrolls, etc. posits the objection that if Luke had written the genitive as a plain noun phrase (”of Quirinius the governor”), the “before” reading would be grammatically open — but the participle and the word order of the way it was actually written shut this up. Fitzmyer is a Jesuit, and while it is true that I’ve spent quite a lot of time exposing the Jesuits, his point is genuinely valid and independent of his associations — because the grammar just doesn’t work. That’s the real issue. Daniel Wallace calls the reading “almost impossible” and Fitzmyer calls it a “last-ditch solution.”
One thing we might be able to say though is that Luke’s intent to distinguish this census from the notorious 6 AD one is entirely plausible, since he explicitly references that later census in Acts 5:37. He obviously knew there was more than one relevant registration. It’s just that “before” is not what the Greek he actually wrote conveys. Either way, keep this part in mind because it will be useful later.
OPTION 2: QUIRINIUS GOVERNING IN A LESSER ROLE
The next argument also relies on re-framing the translation, by positing that “governing” (hēgemoneuontos/ἡγεμονεύω) has a semantic range that can include lesser roles, not only complete or absolute roles.
However this is, too, is problematic because in the next chapter (Luke 3) Luke describes Pilate with the same exact participle: "Pontius Pilate governing Judea." The verb here for “governing” is the same verb and same construction, for a man holding the full office of governor. So “hēgemoneuontos” being a general verb for "exercising authority" is true, but it does not by itself signal a lesser role automatically — Luke uses it for a formal governor too. As a result the grammar is neutral on rank and we need more context.
This pushes us to search for extra-biblical sources to qualify this one way or another. Justin Martyr, writing in Rome ~150 AD, calls Quirinius the “procurator” (ἐπίτροπος) of Judea and invites readers to check the census archives — which does point to Quirinius holding some Roman administrative role tied to a Judean registration, and a role below a full legate. However it’s also important that “procurator” is Justin’s own word, possibly used loosely. That means it’s not a perfect fit for some kind of authority that could conduct a census. At best it’s suggestive, but lightly.
Then there’s the Homonadensian War (roughly 12–2 BC). This is a Roman military campaign against a rebellious tribe, and Quirinius held a significant military command in that period. This would be a natural place to locate an early Quirinius authority in the East. However this too has problems, first and foremost that the war was fought in the Taurus mountains of Asia Minor (Galatia/Cilicia), not Syria or Judea, and the resident legates of Syria for that whole window are well documented — Titius, then Saturninus (~9–6 BC), then Varus (~7/6–4 BC).
So to put Quirinius with Syrian/Judean census authority around 3 BC, you have to posit an additional, undocumented special command layered over or beside the known legate of the time. That’s not totally impossible — Rome did dispatch second imperial legates for wars and special tasks — but it is a historical reconstruction resting on an argument from silence, so it’s not terribly strong either. At least not on face value, but we will come back to it later.
OPTION 3: TWO GOVERNORSHIPS, THE LAPIS TIBURTINUS
The words “lapis tiburtinus” mean "stone of Tibur" (modern-day Tivoli), which later corrupted into the word "travertine". This stone inscription (found near Tivoli in 1764) describes a Roman official who governed Asia, won triumphal honors and was legate of Syria for a second time. Many scholars attribute this to Quirinius, however as with everything presented thus far — there are (surprise surprise) various problems.
The text (translated) of the inscription is reproduced below:
“The honorand subdued a king and brought a kingdom into the power of Augustus, received two thanksgivings (supplicationes) and triumphal ornaments, was proconsul of Asia, and as propraetorian legate of the deified Augustus obtained [iterum, a second time] Syria and Phoenicia.”
Firstly, the name is broken off. It doesn’t say “Quirinius” — his identification is an inference based on the events reported. For roughly a century and a half after the 1764 discovery, most scholars did identify the honorand mentioned here as Quirinius — that’s the Mommsen/Ramsay tradition most apologetics sources draw on. But modern prosopographers have put forward serious alternatives to the traditional view. For reference, prosopography is a historical research method that investigates a group of people by collectively studying their shared background characteristics, careers, and relationships rather than focusing on a single biography; think of it like a branch of anthropology.
The two main ones are L. Calpurnius Piso (was consul around 15 BC), championed by Ronald Syme, the most formidable Augustan prosopographer of the 20th century, and C. Sentius Saturninus (consul around 19 BC), argued by Nikos Kokkinos. Kokkinos makes a detailed case that Saturninus is the best fit and even calls Piso “really a non-starter.” The reason Saturninus keeps gaining ground is that his documented career lines up unusually well with the inscription: he was proconsul of Asia, legate of Syria roughly 9–6 BC, then commander in Germany AD 3–6 with the requisite honors — practically satisfying every line of the stone.
So it’s not that there’s one obvious rival; it’s that at least three men’s careers can be draped over a broken inscription — with the one we want (Quirinius) not even being a top candidate.
One specific point that actually hurts Quirinius here is that the honorand in this inscription subdued a rex, which is Latin for a king. Quirinius’s famous victory was over the Homonadenses, a tribe in the Taurus — not a kingdom with a king in the way the Thracian war (Piso’s) produced one. That single word is part of why Syme leaned with Piso, but in either case it’s one less match for Quirinius.
Another thing to consider is that a major Roman historian, A.N. Sherwin-White, rejected the association of Quirinius. Sherwin-White’s problem was geographical and prosopographical, not grammatical. The Homonadensian war was fought in the Taurus mountains, so if Quirinius held a governorship during that campaign it would have been Galatia, not Syria. And the Syrian governor list for the relevant window is already filled: Sentius Saturninus preceded Varus, serving roughly 9–6 BC. Between Saturninus and Varus there’s simply no vacancy to slot a Quirinian Syrian term before 4 BC. So on Sherwin-White’s reading Quirinius held Syria exactly once — in AD 6 — which means a “twice in Syria” honorand can’t be him.
It’s important here to specify that Sherwin-White was not a Luke-debunker. He defended Luke’s general accuracy and even thought it conceivable Quirinius held some earlier office — but he concluded the census Luke describes can only be the AD 6 one, which is what he saw as inconsistent with the reference to Herod.
Another issue is with the Latin itself, specifically the “second time” modifier before Syria and Phoenicia (“leg. iterum”). The problem is that it is ambiguous. It can be read as “legate of Syria a second time” or as “again a legate [and this time] of Syria,” which would not require two Syrian terms at all. Out of the two, the more defensible reading actually attaches iterum to Syria. Sherwin-White marshaled other inscriptional parallels to argue that in this text iterum modifies “Syria,” not the legateship in general — so whoever the man was, he governed Syria twice.
That sounds like it should help the case, but as we just established through Sherwin-White’s testimony — it actually doesn’t help Quirinius at all because the whole problem is that Quirinius’s career has no room for two Syrian terms. The reading that would help Quirinius is the weaker one — “again a legate, this time over Syria,” where his first legateship was elsewhere (Galatia, for example) and Syria is the AD 6 posting. But that reading gives him no early Syrian term at all, so it does nothing for a nativity-era census either. Either way the grammar doesn’t get much mileage. The debate is settled on prosopography, not on iterum.
But now what arguments do people actually make for Quirinius being the one on the stone inscription? The answer primarily lies on matching the general shape of Quirinius’ career and tradition. Mommsen and Ramsay, prominent 19th-century classical scholars and historians who corresponded and collaborated on ancient epigraphy and Roman history, both backed him as the candidate — and the stone was found near where Varus had a villa at Tibur, placing it in that circle. Quirinius’s career genuinely does include a consulship, a major eastern victory with triumphal honors, and a secure Syrian legateship in AD 6, so at least one of the two required Syrian terms is certain; and Tacitus (Annals 3.48) does record Quirinius’s military distinction, including the Homonadensian campaign.
Therefore the pro-Quirinius case is “his profile fits the broad shape, and he’s the traditional reading.” The anti-Quirinius case is “other men fit the specific lines better, and his career can’t accommodate two Syrian governorships.” The latter is currently winning among specialists, and for good reason — because the traditional case just doesn’t hold up to rigorous scrutiny and modern evidence.
An important note here is that Ernest Martin, the primary scholar who questioned the 4 BC death date for Herod (and which I cite in my own defense of a 1 BC death date, 3 BC birth date for Jesus), argued in an appendix to The Star of Bethlehem that the Lapis Tiburtinus refers to Varus, not Quirinius. So for me this is yet another problem to reconcile in trying to place Quirinius in that timeframe using this inscription, hence I reject the traditional view here just like Martin and Sherwin-White. The stone inscription is not Quirinius, and that at least eliminates option 3 (along with option 1) from our pool.
YET ANOTHER PROBLEM: THE CLIENT KINGDOM
There's yet another difficulty none of the three options dissolves, and that is the following. Under Herod, Judea was a client kingdom, not a Roman province. Rome generally did not run direct censuses in client kingdoms — Herod raised and remitted his own taxes. Direct Roman registration of Judea is exactly what happened in 6 AD because that's when annexation occurred and it's why it triggered a revolt (which again, Luke documents in Acts 5). A Roman census inside Herod's autonomous kingdom around 3 BC is therefore historically anomalous, and there's no independent attestation of one, nor of any Augustan census requiring people to travel to ancestral towns. This is probably the strongest leveraging point for the skeptic, and it’s genuinely a challenge to reconcile — but not impossible as we will soon discover.
WHAT I MAKE OF ALL OF THIS
I said at the beginning of this article that our journey would require nuance and carefulness. On one hand, I believe there is sufficient evidence that is very much in favor of the fact that Quirinius could not possibly have been in Judea governing and orchestrating a Roman census in 3 BC or even 1 BC. That evidence is:
Judea was a client kingdom run by Herod (at that time) with its autonomous taxation and census procedures
The Greek for “first” is very much “first” and not possible as “before”
The Greek for “governing” gives no preference to any kind of special authority or reign, and Luke uses it to refer to Pilate’s full prefecture in the next chapter
The Lapis Tiburnitus has stronger candidates, like Saturninus
There is no window of time given existing records when Quirinius could have governed Judea and executed an official census. Note that what I mean here is governing Judea in an official capacity. What I will present here is an unofficial administrative and temporary role prior to this official capacity, which is what I believe reconciles this problem.
Either way, these are strong objections and there is no simple black and white verse or single piece of evidence we can point to in this case to refute them easily and immediately. That doesn’t mean we lack evidence, though. What I lean on to balance these problems is the following:
The birth date of Christ being 3 BC is very defensible from a variety of witnesses (Daniel’s 70 weeks, Revelation 12 astronomy, Luke’s timestamps, even the Talmud recording 40 years of the day of atonement omens changing before the judgment).
Luke, in general, is highly meticulous and careful with details. He is obviously aware of the 6 AD census, as he documents a historical fact that happened as a result of it (the Jewish revolt). He also documents many places and names in the book of Acts, and in general he is perhaps the most detailed of the 4 gospels in his writing style. Mark is very short and to the point, Matthew relays a narrative style and John is very intimate and personal. Luke is the historian, and that consistency earns him the benefit of the doubt in my mind.
The Bible is God’s word and as such is a supernatural text. There are ample proofs for its legitimacy as a historical document and also as divine revelation that I am not going to open here. That being said, the Bible is also translated by mortal men and things get lost in translation. Sometimes even when the translation is good enough — things still get lost because of absent context; context that was probably present for the author’s audience at the time but is simply not there for us two millennia later. One example of this is Paul’s “prince of the power of the air” (Ephesians 2:2). Many debate what this is, and everyone has theories ranging from plausible to exotic, yet there simply is not enough context to say decidedly what Paul meant — and God for whatever reason has left it that way.
This particular issue is not a life-changing or dogmatic issue. We are not dealing with the deity of Christ or whether salvation is one way or another. In other words, there is literally zero doctrinal impact from this verse. That doesn’t mean we should just gloss over it and let the skeptics have a heyday — but it does mean not giving them the farm and letting them discredit the entire Bible through cherry-picked accusations, either.
With all of this in mind, I lean toward option 2 as the most plausible — that Luke is describing a nuanced political reality of his time that, for his audience, the context would have been clear. There are many reasons for this, but I will preface them by saying that Luke already uses this nuanced pattern of reporting in his meticulous timestamps for Jesus’ arrival in Luke 3, specifically the 15th year of Tiberius and the “high priesthood of Annas and Caiaphas.” I’ve discussed the intricacies of these socio-political realities that Luke is reporting because in both these cases there are overlapping situations. For Tiberius, he was granted co-regency in AD 13 with Augustus, and I believe Luke’s 15 year count starts with that initial point (hence 27 AD for Jesus’ baptism when He was “about thirty”). It’s unconventional to start the ruling period from a co-imperium grant, but Luke is reporting things “as they actually are” not as the proverbial mainstream media would have reported them.
The same goes for the priesthood of Annas and Caiaphas. Annas was the father-in-law of Caiaphas. Though Annas was deposed by Roman authorities, he retained immense informal power, working in the shadows politically while Caiaphas held the official ruling office during the trial and crucifixion of Jesus. The fact Luke ascribes the priesthood to both these men shows you several things:
He is well aware of the political nuances of his time
He reports these political nuances, not just mainstream records
His audience would have picked these things up, otherwise he wouldn’t have written them this way
His relaying of the priesthood this way favors the counting of the 15 years of Tiberius from the co-regency, not from the actual beginning of Tiberius’ solo reign (and this also aligns perfectly with a 3 BC birthdate for Christ and Luke saying Jesus was “about thirty”)
So Luke’s style in chapter 3 reveals that he is extremely careful, and also takes a deliberate and pragmatic approach to reporting certain timestamps based on the ground-level socio-political realities of the time, not necessarily the official books or records that are written later of such events. This is important, because there are plenty of reasons to believe he is doing exactly this with Quirinius and the first “registration” (note: not census, we’ll come back to this) around Jesus’ birth.
THE EVIDENCE FOR QUIRINIUS: AEMILIUS SECUNDUS
The key to making option 2 defensible is to stop defending the wrong thing. In other words, the moment we frame this as "Rome ran a direct tax census inside Herod's independent kingdom" — we have walked onto the skeptic's strongest ground, because the mainstream objection (this is scholar Raymond Brown's formulation) is precisely that there was no single empire-wide census under Augustus, that Rome did not directly tax client kingdoms, that no census required travel to an ancestral town, and that a Judean census wouldn't have touched Joseph anyway if he was in Galilee under a different ruler.
So to accept the frame "direct Roman taxation of Herod's realm," is to get stuck. That’s why this requires nuance, and that nuance is to read the word for “registration” (apographē) as not a tax census at all, but an enrollment of some kind. Why this is a powerful explanation is because there are historical precedents for exactly this.
The Aemilius Secundus inscription (the Lapis Venetus) records that Quintus Aemilius Secundus, serving under Quirinius, carried out a census of Apamea — an autonomous city-state that minted its own coins — by Quirinius's order. Separately, Tacitus (Annals 6.41) reports that around AD 36 the Clitae, a tribe subject to the client king Archelaus, were compelled to render an account of their revenue "in the Roman fashion" and submit to tribute. Together these show that Roman-style registration reached both autonomous cities and client-ruled territory.
Paul Maier's list of surviving Augustan provincial registrations spans Gaul, Sicily, Cilicia, Cyrene, and Egypt, and he explicitly includes client kingdoms such as Herod's in the category of territories where such registrations occurred. And Rome plainly had fiscal visibility into Herod's kingdom: when Archelaus succeeded Herod, he handed Augustus the accounts of Herod's money and the kingdom's annual income (Antiquities 17.228–229).
All of this is significant because the claim that "Rome would never register a client kingdom" is too strong, and that leaves the door open for more evidence to build the case of what is actually being reported by Luke, which I believe to be a special kind of enrollment event.
THE EVIDENCE FOR QUIRINIUS: JOSEPHUS
Now, with that the next thing to consider is that Josephus (Antiquities 17.41–42) records that late in Herod's reign the whole nation gave assurance of goodwill to Caesar and to Herod's government (an oath of allegiance basically) — which some 6,000 Pharisees refused, and were fined for. The telling detail is that Josephus and Herod knew the exact number who refused (over 6,000), which implies an enrollment or count accompanied the oath. You can't fine a defined 6,000 without a register or list of some kind that keeps track of names, addresses, property, etc. This maps onto a documented empire-wide phenomenon: around 3 BC there is considerable evidence that everyone in the empire was required to swear allegiance to Augustus as "Father of the Country" (Pater Patriae), so Luke's “registration” may have been an enrollment for that oath rather than a standard taxation census.
There's even a dated external parallel, via an inscription from Paphlagonia, firmly dated to 3 BC, that records an oath of obedience to Augustus sworn not only by resident Romans but by "all the people" of the region (noted by Thorley and by Syme). And the Breviarium Totius Imperii — Augustus's empire-wide summary of resources, displayed in Rome in 2 BC — has been argued to mark the completion of a general Augustan empire-wide registration. One qualifying caveat is that the displaying of this report that early (2 BC) is argued as a minority position, for example by Béatrice Le Teuff, a French ancient historian and doctor in ancient history specializing in Roman provincial administration, land registries, and imperial censuses. It’s not the majority position, as the majority consider the Breviarium read out in the Roman Senate on Augustus’ death in 14 AD (Tacitus Annals 1.11). Regardless, Augustus did have a general, empire-wide accounting drive and this is the main point. If it was also displayed around the time of 2 BC then that only strengthens the case.
Where this dovetails nicely with Herod’s 1 BC death date (and Jesus being born in 3 BC) is that this oath-registration reading only works if Herod was still alive in 3–2 BC. On the standard 4 BC death timeline, a 2 BC oath falls after Herod is gone and can’t be the nativity census in his reign. But I reject 4 BC as I have documented elsewhere, and 1 BC is exactly what lets the Pater Patriae registration sit inside Herod’s reign and line up with all of these events. So unlike the Lapis Tiburtinus — which creates tension with Ernest Martin (the scholar who argues for the 1 BC Herod death date, but also denies the stone is about Quirinius) — this piece is mutually reinforcing with the rest of the framework.
One caveat is that Josephus calls it an oath, not explicitly a registration. In other words, Luke and Josephus use different names, and this is a reasonable inference. An oath doesn’t obviously require anyone to travel to an ancestral town either. It is true that the Egyptian census papyri show registration “by household of origin” was a real provincial practice in the Roman Empire — but that’s Egypt, not Judea, so there’s that.
But the office that supervised the Judean client kingdom was the legateship of Syria. Recall that Sentius Saturninus was legate of Syria, and he presided over the trial of Herod's own sons at Berytus; Varus, the next legate, intervened militarily in Judea around Herod's death; Augustus had to ratify Herod's will and personally approve the execution of his sons; and when Herod fell out of favor over the Nabatean affair, Augustus warned that he would henceforth treat him not as a friend but "as a subject."
So Herod's "independence" operated under continuous Roman regional oversight, and the instrument of that oversight was whoever held authority over Syria. Yes, Judea wasn’t officially annexed until 6 AD, but again — there were many socio-political realities operating in a field of “grey” well before then because of these power dynamics. Basically, Herod’s client kingdom wasn’t this walled off, autonomous area that just happened to pay Caesar taxes ever year. It was being micromanaged for a variety of reasons: Herod’s falling out of favor, the Jews in general being a rebellious bunch, the political importance of the area, etc.
All of this sets the precedent that lets Luke write "when Quirinius was governing Syria" (recall the Greek word there is the same participle Luke uses for Pilate, so it doesn't by itself signal rank) as the dating point for a registration carried out in Herod's realm — not "Quirinius the tax-collector of Judea," but "during the period when Quirinius held the relevant command over Syria to help orchestrate this registration"; which itself, again, is not the 6 AD census (or any tax census for that matter) but rather having to do with the oath enrollment ledger. The point to be aware here is that Quirinius's holding such a command around 3 BC is inferred from his career, not independently documented in any records or archaeological evidence. But we wouldn’t expect such a thing anyway, because this particular role wasn’t a significant or “solo” enterprise like his official prefecture in 6 AD. Rather this was more of a temporary managerial or “middle man” position in the interim of all of these power changes.
THE EVIDENCE FOR QUIRINIUS: JUSTIN MARTYR
The next point to consider is a return to something previously discussed: Justin Martyr’s reference to Quirinius as the “procurator.” The Greek of his First Apology, Chapter 34 reads that Jesus was born in Bethlehem “as you can learn from the census-lists made under Quirinius, your first procurator (ἐπίτροπος) in Judaea”.
First thing to notice is that the word signifies a financial/administrative agent, not a legate or governor. Michael Kruger makes the point that Justin calls Quirinius “procurator” rather than “governor,” which some read as a memory that his census role (at least the first time, when Luke was reporting) was administrative rather than the full legateship. This is the argument scholar Sabine Huebner uses to argue Quirinius was procurator at the time of Luke’s census and only later became governor. That’s useful and suggestive, but again it’s an inference. It also is a double-edged blade, because Quirinius was never technically procurator or prefect of Judea at all. He was legate of Syria; the first Roman prefect of Judea was Coponius in AD 6. So this cuts both ways, and Justin’s title is loose regardless of date and may be utilizing the same colloquial, socio-political and pragmatic style that Luke is using. In other words, Justin Martyr wasn’t documenting some official title given to Quirinius, but rather a temporary administrative position he was participating in due to the established political dynamics of the area and time.
So we can’t lean hard on a literal, official interpretation of Justin Martyr’s statement, because the office of a Roman administrator in Judea didn’t exist until the 6 AD annexation. The neutral way to read this is that Justin simply was thinking of Quirinius as a predecessor of Pontius Pilate — i.e., placing him in the line of Judean provincial administrators, which is the post-6 AD world. Justin is a mid-second-century apologist working from Luke; he never claims to have inspected the lists himself, only that the emperors could. So his real evidentiary value is narrow, but it’s still useful. Justin’s testimony shows the “nativity-under-Quirinius” tradition was firm and early in the Church, and that Christians believed checkable records existed. He is not an independent witness that Quirinius held office before 6 AD, and the “first procurator” phrase can’t support that claim on a literal level. It is just additional secondary evidence along with everything else that further helps our inference.
THE EVIDENCE FOR QUIRINIUS: TERTULLIAN
Inevitable in any serious discussion of this topic, especially when invoking patristic sources, is to point to Tertullian. He says the nativity-era census in Judea was taken under Sentius Saturninus (Against Marcion 4.19). Recall that Saturninus was legate of Syria roughly 9–6 BC, which fits a census in Herod’s reign, but this only works if you push Jesus’ birth date to 5 or 6 BC, and Herod’s death date to 4 BC, which is the mainstream evangelical (wrong) position. It’s internally coherent, at least on face value, because Saturninus (or the Saturninus-to-Varus transition around 7-6 BC) was the resident legate handling internal affairs with Quirinius attached to Syria’s military/external affairs. Quirinius was later famous for the 6 AD census, so Luke pegs the earlier event to the name everyone associated with “the census.” That’s coherent, but it assumes Tertullian is correct and it assumes Herod died in 4 BC, not 1 BC. One of these (the latter) is not true, so it is unlikely that Tertullian was accurate if we interpret Tertullian in the narrow sense of the census happening (and completing) during Saturninus’ rule.
But that may not be the case, and it’s worth documenting even if it only serves to remove Tertullian as a witness. First, you are dealing with the ancient world. They did not have the internet or cell phones or electricity. Censuses were generally multi-year affairs (the Gallic census dragged on for decades), so one could argue the registration was decreed under Saturninus around 6 BC and the local enrollment that brought Joseph to Bethlehem happened years later, near 3 BC. But Tertullian says the census was taken under Saturninus, implying completion, not initiation, so the only way Tertullian could be rescued is if one could argue that he was conveying the inception of the event rather than the completion of this event. That may be plausible, but in either case I don’t think Tertullian’s testimony is absolutely necessary. I’m inclined to believe he was describing the same thing as Justin Martyr but from a different perspective (the “inception” hypothesis), more than I believe he was just flat out wrong, but either way it’s not necessary for the case.
THE REMAINING THORN: RETURNING HOME
So far we have shaped this out quite a bit, but there is yet one last hurdle and it’s the proverbial thorn in the side for this entire thing: if Luke was reporting an oath registration of some kind, why require people to return to their home towns? This is census/tax behavior, not oath registration behavior. If it was just an oath of fealty, why not just sign there where they were and get it over with? This is the problem, because the harder we lean into a “purely an oath” theory, the more it seems to make this specific detail stand out like a sore thumb.
Thankfully, there are answers here too if we dig deep enough.
The classic example here is the edict of Gaius Vibius Maximus, prefect of Egypt (British Museum Papyrus 904, 104 AD). It is an edict that follows a house-by-house census that was already underway, and he ordered everyone living away from their home districts to return to their own hearths so they can complete the registration and attend to the cultivation of their land. There are two important points here to consider from this precedent.
First, it uses the same Greek word Luke does for “registration” (apographē/ἀπογραφή). Second, scholar Michael Rostovtzeff (20th century emeritus professor of history in Yale University) noted that this “return to the idia” — return to one’s place of origin — is attested in more than one census conducted in the Roman East, not just this single papyrus. That is a crucial precedent, because it answers the objection that “no registration ever required travel back home.” There is a historical precedent for it, and several times at that.
There is a caveat with this, however, and that is that the Egyptian return was to a person’s current domicile and property — the place where they had a household and land to cultivate — not to the town of an ancestor a thousand years back. It does not establish “everyone reported to their ancient tribal seat.” Peter Williams of Tyndale House (a serious conservative scholar, not a skeptic) makes exactly this correction. He points out that Luke never says Rome cared about Jesus’s ancestry a millennium earlier; the census existed to maximize tax revenue, and knowing where people belonged mattered for preventing evasion — so the natural inference is that Joseph had actual holdings or a family domicile in Bethlehem. On this reading, Joseph goes to Bethlehem because Bethlehem was genuinely his town — his family’s base, where he still had economic standing and property — and Luke’s “because he was of the house and lineage of David” explains why that was the family’s town (a Davidic family, rooted in David’s town), not the administrative trigger that summoned him.
So you’re dealing with two realities superimposed and reported to you in a single text, that’s why this is so nuanced. The first reality is the pagan Roman state and the second reality is a Jewish ancestral estate. Rome just cared about taxes and loyalty, the Jews kept track of their inheritances and land and genealogy — Luke is reporting both sides of the transaction so that you (the reader) has the full picture. That does not mean Luke is reporting a census and confusing dates or persons. Rather what is happening has multiple layers of economic and political overlapping realities. Quirinius was temporarily assisting with an important territory that Rome was getting ready to annex due to instability, and a state mandated “registration” of loyalty required a return to where one’s economic base existed for the purposes of enrolling said property to the emperor as a pledge of fealty.
And because Joseph likely had his own inheritance from the Davidic line, it stands to reason that he had property in Bethlehem and therefore he had to return there for the oath registration mandate. Joseph is also called a téktōn (τέκτων). In Greek this is a word that describes not a destitute laborer but a skilled builder or craftsman working in hard materials, a man of a trade and standing, a mason. This cuts against the popular image of a propertyless woodworker peasant. Rather it describes a man with an established household to register, though importantly — it's his Davidic roots in Bethlehem, not his trade, that tie that household to that town.
Now some will argue then, how could Mary give birth in a manger if Joseph had property in Bethlehem? But owning ancestral land does not equate to having a guest room free at peak registration crowding when you’ve got your entire extended family and neighbors all returning to the area. The influx of everyone back to their estate meant some (the young couple, for example) would have to make sacrifices.
So if Bethlehem was Joseph’s real family town, the crowded guest room makes sense — the “kataluma” (lodging place, Luke 2:7) being full points to other relatives also converging on Bethlehem, which is what you’d expect of a big Jewish family with roots there. And the tradition that the family remained in Bethlehem afterward (they’re in a “house” when the Magi arrive) fits a man who actually belonged there. Again, having family ties to a town is not the same as having a spare room during a registration frenzy.
The catch here is that this reading is a property/domicile explanation, and that puts this thesis in tax-census territory. We don’t want to lean on it too heavily, because we return back to the original problem of the tax census. But there is enough of a precedent, and supporting data, that the unique shape of this event is becoming clearer.
A registration administered locally by Herod’s own apparatus among Jews would naturally carry an allegiance dimension (that’s the oath Josephus records, and how anyone knew that 6,000 Pharisees refused) and would naturally be organized the Jewish way — by household, clan, and ancestral town — because that is how Jewish identity and records actually worked, from tribal allotments to priestly courses to the genealogies kept at the Temple. On that picture the “un-Roman” quality of this event isn’t an anomaly, it’s the tell that this wasn’t a direct Roman census at all. It was a Herodian enrollment of the Jewish nation, which is precisely why it could happen in a client kingdom and why it followed genealogical rather than pure property lines. Joseph’s return to Bethlehem then sits at the intersection of these realities quite neatly, because it was both his family’s town and his genealogical seat — so we don’t have to choose between “he had ties there” and “he belonged there by descent.” Both are valid, and given the socio-political context just described, it all fits very consistently with the rest of the evidence.
All that being said, the specific claim that a Herodian enrollment (and assisted by Quirinius as some kind of interim administrator) was organized by ancestral town is a reconstruction, not something we actually have a document for. The Egyptian papyri prove return-to-home registration existed in the Roman East, but they’re a century later, a different province, and about domicile rather than ancestry — so they establish plausibility, not full proof, and the critics are within their rights to say the extrapolation to Herod’s Judea isn’t automatic or guaranteed. The defense is narrow but it exists in a very real way, because a “return-home registration” is documented Roman-era practice, and a locally-run Jewish enrollment would plausibly key on family towns — which for a Davidic family was Bethlehem. That’s all very plausible. Saying that Rome required everyone to trek to the seat of their thousand-year-old ancestors is not, and that’s not what’s on the table here. So contextualizing this in the Judean realities of the time is what makes it different than Egypt or other places described.
CONCLUSION
This is quite the historical rabbit hole, and it reminds us that when it comes to history things are not always black and white. I believe these are moments of faith, because certain things are lost to the sands of time. For example, we know Daniel is a highly historical book — yet the events of Daniel 11 (at least the first half of that chapter) are highly nebulous and debatable because we simply don’t have all the necessary concrete records. That doesn’t mean Daniel is discredited or reporting things that aren’t true, it just means that there is an element of history in Bible study that is sometimes grey. And maybe the events in question are not relevant for our generation, hence why God did not preserve them with full clarity.
The good news is that all the key events that we do need to know about are extremely well documented. We have great evidence for all the prophecies of Daniel (Daniel 2, 7, 8 and 9) and we have robust evidence for the historicity of Jesus, the crucifixion and the resurrection. What matters is that God has preserved His word and preserved enough evidence for us to believe the important pillars of faith, while leaving some things as points of struggle and wrestling so that we simultaneously sharpen our knowledge while growing in faith. There are many things in scripture I don’t have an explanation or answer to, but that doesn’t mean I stop believing the Bible. I know the Bible is true, so I trust there is always an explanation that will be revealed in time — even if that time is on the other side of eternity.
With that in mind, I believe that Luke (and it seems the people of that time in a general sense, like Tertullian or Justin Martyr) had a way of reporting certain facts that were very “street smart” and pragmatic, not just “book smart.” Luke for certain, because he establishes that pattern with Annas and Caiaphas and Tiberius. So what is being reported is not the census of 6 AD, which he specifically pinpoints later in Acts 5 (by naming the revolutionary who led it specifically) — but rather an event that obviously happened in a period where Judea was approaching transition politically. Herod was losing favor (the Nabatean debacle), Judea was going to be annexed, Quirinius was placed as an interim (but relatively undocumented/unofficial) administrator for a Roman oath of allegiance that was declared by the Emperor, and because Herod was the ruler of the client kingdom — the oath in question for Judea had both a Jewish lineage and a socio-economic frame. People returned to their homes for economic purposes, and it also happened to be that Joseph was from Bethlehem by being of David’s line.
Now, it’s true that a pure loyalty oath does not explain a return-to-home at all. We can see this directly. The Paphlagonia oath of 3 BC was sworn at the altars of Augustus in the temples of each district — people swore locally, wherever they were. If sentiment were the whole point of the oath then you’d walk to the nearest altar, swear, sign, and go home. The oath has a location element, but it’s your current district, not your estate. So the moment Luke has people traveling to their economic home base, he’s describing something a loyalty oath by itself would never require. That detail is telling you the registration had a second, fiscal character — which is where the return-to-home comes in.
By having people return to their estates, the registration anchors each person to property and jurisdiction. Roman registration in the East was not a signature drive; it was, in its own words, the “house-by-house” registration. The Vibius Maximus edict opens by referring to the census “by household” and orders people back to their own hearths precisely so they can complete the registration and attend to the cultivation of their land.
Note that both registration and land are tied together in a single edict. The unit being counted is not the individual but the household situated on its holding. That’s the whole answer to the pure oath objection. “Can’t you just sign your name?” No, you can’t. Not in this situation. The ledger’s index was the house, and a house has a place. An individual signature detached from a household and a location is worthless to a governing power, because what Augustus wanted was not a pile of names but a governable map: this household, of this many souls, on this estate, in this district, under this authority, producing this much. The Breviarium Totius Imperii — his empire-wide summary of resources (possibly displayed in Rome around 2 BC) — was exactly that kind of accounting, and the Res Gestae boasts of census figures as a measure of his dominion. “I acknowledge Caesar” tells him nothing he can rule with. “This accounted-for household at this estate with XYZ property that acknowledges Caesar” tells him everything he needed to know, especially in the run-up to the annexation of this volatile territory.
So that’s why this fealty oath registration had a fiscal component. The house or estate was the thing being pledged in loyalty, with the people attached to it as members. They weren’t delivering empty signatures, they were being enrolled into a specific household-on-its-land index, and the household can only be enrolled where it sits. People like Joseph traveled all around the territory for whatever reason, but their estate is what gets invoked in an empire-wide edict like the one Luke is describing.
Another important thing to consider is that under Augustus these were not two competing purposes but one fused act. In the Pater Patriae decree, to be enrolled was simultaneously to be entered in the imperial account and to acknowledge the man the account belonged to. The allegiance and the accounting are the same thing — submission expressed as being counted. So Josephus’s oath and Luke’s registration aren’t rival events; they are the two faces of the same coin that these authors were documenting. The oath is why refusers could be identified and fined (you can’t fine an anonymous holdout — you can only fine a named household on the rolls, and you can also impose liens and seizures on named property); so the household-and-estate structure is why anyone had to move.
And because Herod was administering an enrollment of his own subjects — counting his households and holdings, in the Jewish manner by family and ancestral town, with an acknowledgment of Caesar attached — there’s no direct Roman taxation of the kingdom at all. It’s a client king taking stock of his realm and demonstrating a loyal, well-ordered population to an emperor he’d fallen out with. That Herod’s kingdom kept exactly this kind of accounting is not speculative: when Archelaus succeeded him, he handed Augustus the accounts of Herod’s money and the kingdom’s annual income. The estate data in Luke is Herod’s own ledger of his own realm, bundled with the loyalty gesture to satisfy the Emperor’s decree.
All of this means that probably the best way to think of this “registration” is less of a “loyalty oath” and more of a “household enrollment” that carried a loyalty dimension given the political overlap and dynamics of that time and place. It wasn’t “just an oath” because it was much more than that. Oaths don’t need travel, but household enrollments (especially under the pressure of a client king obeying orders to show fealty) are completely reasonable. This bridges travel, household structure, the ancestral town, and the oath all in one act as complementary realities rather than competing narratives.
And Quirinius? His role in this affair was as an interim administrator to oversee Herod’s efforts and make sure everything was in order — a position that was not some official position but rather a temporary “assistant to the regional manager” (much like Dwight Schrute in The Office) type of position working with both Herod and Varus in some way. Because Quirinius was well known when the 6 AD revolt happened, his later fame is brought backward in time as a way to identify him. As in “that” Quirinius, the same one that was part of the 6 AD census and revolt.
This I believe is the most defensible conclusion based on the evidence provided and based on the assertion that Jesus was born in 3 BC and Herod died in 1 BC. These nuances are inferred from a variety of sources both biblical and extra-biblical, they are not overtly stated by any one single document or verse. However this issue has no doctrinal impact, so we do not need extensive biblical documentation. We have more than enough for a reasonable and balanced explanation, and we have enough from Luke to give him the benefit of the doubt that he wasn’t being clumsy or confused — but rather incredibly pragmatic and detailed with the highly complex political dynamics of his time.
For a treatment of these and other theological topics in detail, please see my book “The Seven Transformations: Finding the Narrow Road by Pursuing a Sovereign Christ.”
