This is the strange theory advanced by John Loftus in chapter 4 of God or Godless?.

I call it ‘strange’ because I had not come across it before.  I hastily looked up the passage he cites (Exodus 22:29-30) to see what I had been missing all these years.  We read:

You shall not delay to offer from the fulness of your harvest and from the outflow of your presses.  The first-born of your sons you shall give to me.  You shall do likewise with your oxen and with your sheep.

Taking this in context, says John, it means that God required firstborn sons to be literally sacrificed.  It is only much later in OT times that God is said to change his mind, and to outlaw child sacrifice.

This is extraordinarily inept.  As Exodus 13:13-15 has already made clear, first-born males of humans (and certain animals) were not slaughtered; they were redeemed – they were bought back from God by the payment of a price.

This rejection of child sacrifice goes right back to Abraham, who demonstrated his willingness to sacrifice his son Isaac but found that a substitute had been divinely provided.  John cites varous examples of child sacrifice in the OT, but these can easily be shown to be beside the point, since being regarded by the biblical writers themselves as abhorrent as we ourselves find them.  To take just one example – the first mentioned by John – Jepthah’s action is sacrificing his daughter (Judges 11:29-40) is clearly regarded by the narrator as an horrendous act carried out as a result of foolish oath.  The rather obvious point that John seems not to understand is that Scripture does not necessarily approve everything that it records.  Far from it.

Turning to Randal’s statement, I find that he not even talking about the same problem as John.  (That, by the way, if one of the weaknesses of the debate format of this book: each writes his opening statement without having seen the opening statement of the other).  Noting that the OT explicitly repudiates child sacrifice (Leviticus 18:21; Ezekiel 23:37) he assumes that the subject to be debated is killing of the children of the Amalekites (for example; see 1 Samual 15:3) by the Israelites.  That’s a stiff enough problem in its own right, but not what John means by ‘child sacrifice’.  Randal then proceeds to suggest that the texts that indicate that God commanded these ‘total’ killings might be actually mistaken.  But that’s not dealing with the problem, it’s avoiding it.

I have a little sympathy with Randal on this.  He wants to say, in effect, “We don’t know what to make of these problem passages.”  I’m not opposed in principle, to saying, “I don’t know.”  I just think he might have tried harder on this occasion.

So, my verdict is, ‘a plague on both your houses’ on this one.  It’s a goalless draw.  John has completely misunderstood the teaching of the Old Testament regarding child sacrifice.  And Randal has, in his attempt to deal with a rather different problem, simply ducked the issue.