First off, when a birth defect does happen how can you even be sure Substance X caused it? Sometimes it's obvious, sometimes it's not.

Secondly, how do you go about tracking that? Certainly we don't have a big database for the genome of every baby born.

Third, what about substances that don't do acute harm, like lead? Lead poisoning is slow and deteriorates your mind. We know about it now, but for hypothetical new substance X how do we measure who got the poisoning and how much they have (do we take measurements every few years and plot a graph, adjusting the cost over time?)

Fourth, how do we measure esoteric costs? For example this substance causes loss of intelligence. That could translate into lower paying jobs over a lifetime - could. Do we compare average intelligence income to the income of these individuals and then the difference is the cost? What if some of the individuals could have been geniuses, how do we tell? In general, how do we measure against things that have not happened in our reality?

Consider that artificial pricing always creates the Cobra Effect: unintended consequences abound that eventually more than undermine the original intent. This is how George Soros legendarily broke the Bank of England - by taking advantage of artificial pricing (in particular, exchange rates).

Furthermore, healthy societies enforce moral lines through criminal and civil law, not merely through fines and price manipulation.