I guess the population records were wrong, then.
The alternative explaination is that the population records are correct and the assumption that the growth rate of the human population has been completely uniform over the past 6000+ years is incorrect.
To determine the more likely, have a look at how many estimates of population number and birth/death rates would have to be wrong for the former to carry any weight.
For example, the best rebuttal of the population argument I've seen is in Micheal Shermer's "Why people believe wierd things", where he points out that if you use the same population extrapolation you would end up with Egypt being populated by a ridiculously small population at a time where the historical records clearly show a flourishing civilisation.