One of the thorniest historical problems in the Gospel of Luke is the census that Luke reports in chapter two, verses 1-2. Here is the wording in the NIV: “In those days Caesar Augustus issued a decree that a census should be taken of the entire Roman world. (This was the first census that took place while Quirinius was governor of Syria.)”

Here is the problem. Extrabiblical sources indicate that Herod the Great died in 4 BC. Biblical texts indicate that Jesus was born before Herod died, between 4-6 BC. The Jewish historian Josephus wrote that Quirinius became governor of Syria in AD 6 and ordered a census that year. So how could there be a census decreed in 4-6 BC if Quirinius wasn’t governor for ten more years? Did Luke make a mistake?

The first thing to note is that Luke seems to be aware of the census ordered in AD 6, as evidenced by Acts 5:37. Therefore, it is highly unlikely that Luke had his facts mixed up. So how do reconcile the data?

Several suggestions have been made, although none definitively solve the problem. Darrell Bock writes, in The Gospels and Acts (The Holman Apologetics Commentary on the Bible),

One option for solving the problem is to claim that Quirinius was governor twice, with his first governorship occurring prior to the well-known one that was initiated in ad 6 (Ramsay 1898, 174– 96; 1920, 275– 300). Ramsay argued that Quirinius served as governor of Syria from 11/ 10 bc to 8/ 7 bc, as well as in the later and firmly documented period that began in ad 6. Ramsay further argued that prior to the close of his first governorship Quirinius set in motion the census that Luke placed in Palestine in 6 bc.

The basis of Ramsay’s argument for two governorships is his contention that an inscription known as the ‘Lapis Tiburtinus’ refers to Quirinius. There is a noteworthy weakness to this approach: the Lapis Tiburtinus inscription is fragmentary and mentions no names. Thus Ramsay can suggest the inscription pertains to Quirinius only because the portion of the inscription that specifies who is being referenced has been lost. Based as it is on the anonymity of the inscription, Ramsay’s theory is impossible to establish as fact. Possibly he is correct, but there is no way of proving it unless we discover the missing portions of Lapis Tiburtinus.

Bock continues:

In an alternative theory, A. N. Sherwin-White suggested that Quirinius was a legate between Publius Quinctilius Varus and Gaius Caesar from 4 bc to 1 bc (Sherwin-White 1963, 162– 71). He was able to postulate this because this time frame coincides with the only governorship gap we have in the historical records. Thus, though Sherwin-White denied that a census was undertaken in the period of Herod the Great (ended 4 bc), he saw it as possibly legitimate for Luke to make reference to Quirinius in this period.

Bock, however, thinks that Sherwin-White’s hypothesis needs to be modified.

Variants of Sherwin-White’s proposal may offer the best explanation. Given that a Palestinian census under Herod is possible since it fits with Roman census-taking activities known to have occurred in this period, might it be that Varus was the one who began the census, with the full results and taxation emerging only later, under the governorship of Quirinius in the period Sherwin-White proposes (4– 1 bc)? If so, Quirinius’s name was understandably attached to the census since the census’s impact was felt chiefly during his governorship. A variation of this theory is that Quirinius administered the census prior to being governor, later became governor, and thus in a kind of literary collapse of the event was referred to as the governor responsible for the census (Hayles 1974, 29).

Bock offers further hypotheses from lexical-syntactical scholars:

One lexical argument says that prōtē should be translated as ‘earlier’ rather than ‘first.’ This meaning is found in John 1: 15, 30, for instance. If this is a legitimate rendering of prōtē, Luke was saying that the census he has in mind occurred earlier than the Quirinius census known to have taken place in ad 6. This view is possibly correct, but the difficult syntax of Luke 2: 2 leaves the matter unsettled.

A second lexical solution holds that we should translate prōtē as ‘before.’ In this case Luke is saying the census took place before Quirinius’s governorship (Higgins 1969, 200– 201; Nolland 1989, 101– 2). It is questionable whether Luke used prōtē in this unusual sense, but possibly he did (Hoehner 1977, 21– 22).

These alternative renderings of prōtē may not be necessary. If prōtē is best rendered ‘first,’ then Luke was calling this the ‘first’ census that occurred during Quirinius’s governorship. This could indicate that Luke knew of two governorships for Quirinius and two censuses associated with him. The fact that in Acts 5:37 Luke referred to a census best identified as the one that occurred under Quirinius in ad 6 strengthens this possibility. Thus if Luke meant to differentiate between two separate censuses taken by Quirinius, it is best to assume that the census he discussed in chapter 2 of his Gospel is not to be confused with the better-known Quirinius census of ad 6.

Bock concludes:

None of these proposals is guaranteed to be correct. Uncertainty about the succession of governors in Syria prevents us from making anything more than suggestions. The best possibilities are (a) that Varus (during the time of Herod the Great) began the census, but the full effects were not felt until the governorship of Quirinius, (b) that Quirinius served as a high-raking administrator of the census prior to his governorship, and (c) that prōtē is best translated as ‘before.’