Finlan doesn’t think there’s any idea of substitution in the sacrificial rites, though there is in the scapegoat ritual (Paul’s Cultic Atonement Metaphors, 91).
To see a penal substitution in sacrifice, “assumes something that is never stated in Hebrew texts: that the
animal is being somehow punished.” He claims that the contrary is the case: “no abuse is poured on the
sacrificial animal as there is on the scapegoat; it is treated with great care and is
sacrificed to God, while the sin-carrying scapegoat is abused and driven out
to Azazel, something that never happened with sacrificial animals. Beating and
stabbing and spitting would alter the animal from its previously ‘spotless’ condition, rendering it impure. But a sacrificial animal remains spotless, it
brings no impurity into Yahweh’s house. The wilderness, on the other hand, is
an appropriate repository of impurity.”
There are multiple problems with this analysis. First, the evidence for the abuse of the scapegoat is late. There is no hint of it in Leviticus, the canonical prescription for the rite. I don’t doubt that abuse happened on Yom Kippur, but one certainly can’t use those later sources as evidence for the logic of sacrifice and atonement according to Leviticus. Second, the “great care” bestowed on the animal misses a pretty important detail: the sacrificed animal was killed and dismembered (carefully, no doubt) and its blood spread around the sanctuary furniture. This doesn’t prove that the animal is being punished; as Finlan says, that’s never specified in Leviticus. But it does soften the stark distinction Finlan wants to make between sacrifice and scapegoating.