Did Luke Make a Historical Mistake About Jesus’ Birth?

Raymond E. Brown on Chronology, Roman Administration, Bethlehem, and Historical Error in Luke’s Infancy Narrative

One of the most difficult historical problems in the infancy narratives concerns the census described in Luke 2:1–5. For many Christian readers, the story is familiar enough that its historical details are rarely questioned. Caesar Augustus issues a decree, a census takes place while Quirinius governs Syria, and Joseph travels from Nazareth to Bethlehem because he belongs to the house of David. This brings Mary and Joseph to Bethlehem, where Jesus is born.

In The Birth of the Messiah, however, Raymond E. Brown devotes an entire appendix to this problem. What makes Brown’s treatment especially important is that he is not approaching Luke as a hostile critic trying to discredit Christianity. Brown was a major Catholic New Testament scholar, and throughout the book he repeatedly attempts to distinguish legitimate historical criticism from theological judgments about inspiration. Nevertheless, when he examines the evidence surrounding the census, his conclusion is remarkably clear: the chronology of Luke 1 and Luke 2 cannot plausibly be reconciled, there is no convincing historical evidence for the census scenario required by Luke’s narrative, and Luke appears to have made a historical mistake.

Brown begins with the basic chronological problem. Luke 2 connects Jesus’ birth with a census conducted under Publius Sulpicius Quirinius. Historically, Quirinius became governor of Syria in AD 6 and conducted the well known census of Judea in AD 6–7. If Luke 2 stood completely by itself, Brown explains, this would place Jesus’ birth around that period.

But Luke 2 does not stand by itself.

Luke 1:5 places the events surrounding the birth of John the Baptist “in the days of Herod, king of Judea.” Herod the Great died in 4 BC. Luke 1:36 then places Mary’s pregnancy approximately six months after Elizabeth’s. On Luke’s own narrative chronology, Jesus would therefore have been born no later than approximately 3 BC. This corresponds roughly with Matthew’s placement of Jesus’ birth during the reign of Herod the Great.

The problem is obvious. Herod died in 4 BC. The census of Quirinius occurred in AD 6–7. We are dealing with a gap of roughly a decade.

Brown therefore writes:

> “When we compare it with other chronological information that he has given us in 1:5 and 3:1,23, there seems to be an irreconcilable conflict.” Raymond E. Brown, The Birth of the Messiah, p. 547

This is not merely a disagreement between Matthew and Luke. Brown sees a chronological tension inside Luke itself. Luke 1 points toward a birth during the reign of Herod, while Luke 2 connects that birth to an event which historically occurred about ten years later.

Brown considers three broad solutions. One could reinterpret Luke’s reference to Herod so that it no longer refers to Herod the Great. One could attempt to move the census of Quirinius backward into Herod’s reign. Or one could accept that the chronological information in Luke is confused and that reconciliation is neither necessary nor historically possible.

Brown ultimately considers the third option the most plausible.

The first proposed rescue is to argue that Luke 1:5 may have meant Archelaus rather than Herod the Great. Archelaus ruled after Herod’s death and was sometimes associated with kingship. This could potentially move the chronology closer to AD 6.

Brown finds this highly improbable. More importantly, it creates additional problems rather than solving the existing ones. Luke 3:1 and 3:23 place the beginning of Jesus’ ministry in the fifteenth year of Tiberius, approximately AD 27–28, when Jesus was “about thirty years old.” That fits reasonably well with a birth around 4–3 BC. It fits very poorly with a birth in AD 6 or 7.

Thus the internal chronological framework of Luke itself continues to point toward a birth before the census of Quirinius.

Brown then examines a second major question: did Augustus really order the kind of census Luke describes?

Luke 2:1 states that a decree went out from Caesar Augustus that “the whole world” should be registered. Brown acknowledges that Augustus had a strong interest in censuses, taxation, population statistics, and administrative organization. There were censuses of Roman citizens and various provincial registrations during his reign.

But that is not the same thing as evidence for one universal census of the entire Roman Empire at the time Luke requires.

Brown explains that there was no known single census covering the whole Empire in the straightforward modern sense that readers might infer from Luke. Different provinces had different legal statuses and censuses were conducted at different times.

Brown is therefore prepared to allow that Luke may be summarizing Augustan census policy in broad language. But that concession does not solve the larger historical difficulties.

An even more serious problem is the way Luke describes Joseph’s registration.

According to Luke, Joseph travels from Nazareth in Galilee to Bethlehem because he is descended from David. This is essential to the narrative because it explains why a Galilean family happens to be in Bethlehem when Jesus is born.

Historically, however, Brown finds no clear evidence that Roman censuses required people to return to the ancestral hometown of an ancestor who had lived many centuries earlier.

He writes:

> “We have no clear parallel for such a practice.”

p. 549.

Roman taxation normally registered individuals where they lived, owned property, or could actually be assessed. Brown discusses an Egyptian papyrus from AD 104 that is frequently cited in Christian apologetics. It required temporary residents to return to their regular place of residence for registration.

But Brown points out that this does very little to support Luke.

The Egyptian document concerns returning to one’s actual domicile, not travelling to the ancient city of one’s remote ancestors. Joseph, according to Luke, permanently lived in Nazareth. There is no indication that he owned taxable property in Bethlehem.

Brown therefore concludes concerning the Egyptian parallel:

> “It offers little support for sending Joseph from Nazareth where he permanently resided to Bethlehem where clearly he had no property or wealth.” p. 549.

This distinction is crucial. Returning to your legal residence for tax registration is one thing. Travelling to the hometown of an ancestor from roughly a thousand years earlier because you descend from David is something entirely different.

Brown does allow one theoretical possibility. Roman administration sometimes adapted itself to local customs. Perhaps a census in Judea could have taken Jewish tribal or ancestral attachments into consideration.

But possible is not the same as evidenced. Brown finds no historical documentation establishing that this actually happened.

The geographical problem becomes even worse when Galilee is considered.

The census of AD 6–7 was a census of Judea after Archelaus had been deposed and Judea was incorporated into direct Roman administration. Galilee was not part of that province. It remained under Herod Antipas.

Yet Luke’s narrative requires the census to affect Joseph, a resident of Nazareth in Galilee.

Brown therefore notes that the known census of Quirinius does not naturally explain why a Galilean subject of Herod Antipas would have been required to register in Judea.

This leads directly to one of the most common attempts to defend Luke: perhaps Quirinius served as governor of Syria twice and conducted an earlier census during the reign of Herod the Great.

Brown examines this possibility carefully.

There are gaps in our knowledge of the governors of Syria. In principle, therefore, one might try to insert an earlier term of Quirinius into one of those gaps. But Brown argues that what we know about Quirinius’ career makes this unlikely.

He writes:

> “However, from what we know of the relatively well-documented career of Quirinius, it is unlikely that he had an earlier governorship at either of those periods.” p. 550.

Quirinius had a significant and comparatively well documented political and military career. Brown notes his activities in Asia Minor, his military campaign against the Homonadenses, and his later presence in the East as an adviser to Gaius Caesar. Yet there is no clear evidence that he served an earlier term as governor of Syria.

Josephus is especially important here. Josephus discusses the beginning of Quirinius’ Syrian governorship in AD 6 several times.

Brown observes:

> “Josephus, who describes several times the beginning of Quirinius’ legateship in A.D. 6, gives no hint that Quirinius had served previously in that capacity.” p. 550.

Brown also addresses archaeological evidence sometimes invoked in defense of an earlier governorship.

One example is the Lapis Tiburtinus, an inscription concerning an unnamed Roman official who served twice as a legate. Some have identified this person with Quirinius.

Brown is unimpressed:

> “The thesis that this is a reference to Quirinius is a pure guess with as many dissenters as adherents.” p. 551.

Another inscription from Antioch of Pisidia has likewise been used to support the theory. Brown argues that the conclusions drawn from it go considerably beyond what the inscription itself states.

His broader conclusion is straightforward: neither inscription provides convincing evidence for an earlier governorship of Syria by Quirinius.

But even granting such a governorship would not solve everything.

There would still be the question of why Rome would conduct a census for direct taxation inside the kingdom of Herod the Great. Herod was a Roman client king. He administered his own kingdom and collected his own taxes. The famous census of AD 6 took place precisely because Judea had ceased to be ruled as a client kingdom and had come under direct Roman control.

Brown therefore asks why we should assume a Roman census of Herod’s territory before that transition.

Various arguments have been proposed. Augustus once became angry with Herod and threatened to treat him more like a subject. Herod required an oath of allegiance to Augustus. Josephus mentions tax records. Another ancient example shows a client ruler conducting a census in his own territory.

Brown examines these arguments but finds none sufficient to establish the Roman census required by Luke.

The evidence from Josephus points in the opposite direction. When the Roman census finally did occur in AD 6, it provoked enormous controversy. Judas the Galilean led a revolt against it, an event remembered as historically significant and connected with resistance to Roman taxation.

Brown considers Josephus’ complete silence about a comparable Roman census during Herod’s reign significant:

> “The silence of Josephus … is an eloquent argument that there was no Roman census of Palestine before the census under Quirinius in A.D. 6–7.” p. 552.

This also undermines the idea that Jews were already accustomed to Roman censuses before AD 6. Josephus describes the census of Quirinius precisely as a major new development associated with Roman control.

Brown then examines another apologetic proposal: perhaps the text of Luke 2:2 originally referred not to Quirinius but to Saturninus, an earlier governor of Syria.

That would move the census into a chronologically more convenient period.

There is one fundamental problem. There is no manuscript evidence for this reading.

Brown explicitly says:

> “I did not mention this as a textual variant … for there is no manuscript evidence that supports it.” p. 553.

Appeals to Tertullian do not rescue the theory either. Tertullian refers to censuses conducted under Saturninus, but Brown explains that this does not demonstrate that Tertullian possessed a different text of Luke or that Luke originally named Saturninus.

Another attempt argues for a two stage census. Perhaps the registration began under Saturninus or Varus during Herod’s reign and was only completed under Quirinius years later.

Again Brown finds the proposal historically and terminologically weak. Luke does not say that a census begun under one governor was completed under another. Attempts to distinguish separate technical stages using different Greek census terminology do not survive comparison with Luke and Josephus.

Brown remarks that another version of this theory proposes a Jewish census under Herod which Luke later confused with the Roman census under Quirinius. But that solution comes at a price:

> “But this is tantamount to a recognition that Luke was inaccurate.” p. 554.

And this brings Brown to his final historical judgment.

After examining the chronology, Augustus, Roman census procedure, ancestral registration, Galilee, Herod’s political status, Quirinius’ career, Josephus, inscriptions, Tertullian, proposed textual variants, and multi stage census theories, Brown writes:

> “When all is evaluated, the weight of the evidence is strongly against the possibility of reconciling the information in Luke 1 and Luke 2.” p. 554.

That sentence should not be minimized.

Brown is not saying that there is merely a puzzle for which several equally plausible answers exist. He believes the cumulative evidence weighs strongly against harmonization.

He immediately becomes even more explicit:

> “There is no serious reason to believe that there was a Roman census of Palestine under Quirinius during the reign of Herod the Great.”

p. 554.

Brown is willing to allow that Luke 1 may preserve the correct general chronology. Jesus may indeed have been born near the end of Herod the Great’s reign.

The historical mistake, in Brown’s judgment, occurs when Luke connects that birth with the census of Quirinius:

> “But Luke seems to be inaccurate in associating that birth with the one and only census of Judea (not of Galilee) conducted in A.D. 6–7 under Quirinius.” p. 554.

This is the central conclusion of Appendix VII.

Brown does not stop there. He notices that Luke appears to have another chronological problem involving the same historical period in Acts 5:36–37.

In Acts, Gamaliel mentions Theudas and then Judas the Galilean. Yet according to Josephus, the rebellion of Theudas occurred around AD 44–46, long after the time in which Gamaliel’s speech is supposedly taking place. Judas the Galilean, by contrast, belongs to the period of the census in AD 6.

Brown therefore says:

> “Luke also seems to be inaccurate about that census in Acts 5:36.”

Brown follows a proposal that Luke may have confused two major historical disturbances. One was the crisis surrounding the end of Herod’s reign. The other was the annexation of Judea and the census of AD 6. Both were memorable events, and in a society without modern chronological documentation they could eventually become associated incorrectly.

This gives Brown a plausible explanation for the origin of Luke’s mistake without requiring him to assume that Luke deliberately invented history.

But Brown’s final observation is perhaps the most important for understanding how historical criticism approaches the infancy narratives.

Even if Luke got the chronology wrong, the census performs an extremely effective literary and theological function.

Brown writes:

> “Even if Luke was inaccurate on this dating of the census of Quirinius and mistakenly thought that it could have been associated with the birth of Jesus, we must recognize that this association enabled Luke to explain why Joseph and Mary were in Bethlehem when the child was born.” p. 555.

That is critical.

The census gets Joseph and Mary from Nazareth to Bethlehem.

Luke needs Jesus to be born in Bethlehem while simultaneously presenting Nazareth as the family’s normal home. The census becomes the narrative mechanism that moves them there. It also allows Luke to place the birth of Jesus against the vast backdrop of Roman imperial history and the history of Israel.

So Brown does not conclude that Luke’s story is meaningless because its historical chronology fails. He argues instead that the evangelist has used a historical event, the famous census of Quirinius, in a chronologically inaccurate way because it functions effectively within his theological narrative.

That distinction matters.

The problem for a traditional doctrine of strict biblical inerrancy is therefore considerably deeper than merely asking whether a minor date can be adjusted. Brown has examined the standard escape routes and still concludes that the evidence is strongly against reconciliation.

Herod and Quirinius belong to different historical periods. The known census was conducted in AD 6–7. It concerned Judea, not Galilee. There is no convincing evidence that Quirinius governed Syria earlier. The inscriptions frequently invoked to prove such an earlier governorship do not establish it. Josephus gives no indication of an earlier Roman census under Herod. Roman census practice does not provide a clear parallel for requiring Joseph to travel to the city of a remote ancestor. There is no manuscript evidence for replacing Quirinius with Saturninus. The two stage census theories lack adequate historical support. And Luke appears elsewhere in Acts to have confused events associated with this same period.

Brown’s final position is therefore difficult to soften: Luke has apparently made a historical error.

That does not mean Brown rejects Luke’s theological message, Christianity, or even biblical inspiration. He does not. What he rejects is the assumption that inspiration requires every historical and chronological statement in an infancy narrative to function like precise modern historiography.

For readers who approach the Bible with the belief that every historical statement must be factually exact and perfectly harmonizable, Brown’s analysis presents a serious challenge. The difficulty is not created by a skeptical critic imposing an objection from outside the text. It arises from comparing Luke with Luke, Luke with Matthew, and Luke with the external historical evidence.

The final question is therefore not simply, “Can we imagine some scenario in which Luke might still be right?”

Historically, almost anything can be protected by enough hypothetical possibilities.

The more serious question is the one Brown actually investigates: What explanation best fits the evidence we possess?

And Brown’s answer is clear. The weight of that evidence does not support the traditional harmonization. Jesus may have been born during the reign of Herod the Great, but Luke mistakenly associated that birth with the census conducted under Quirinius roughly a decade later. The census then served an important theological and narrative purpose by bringing Joseph and Mary to Bethlehem.

That is a much more significant conclusion than simply saying that “scholars debate the date of Jesus’ birth.” Brown is arguing that at this point the Lucan infancy narrative preserves theology and narrative purpose at the expense of precise historical chronology.

Source: Raymond E. Brown, The Birth of the Messiah: A Commentary on the Infancy Narratives in Matthew and Luke, Appendix VII, “The Census under Quirinius,” pp. 547–555.