People are downvoting, but yes this is accurate. There are confounders between houses with lead and other demographics and contribute to the gap and aren't com

People are downvoting, but yes this is accurate. There are confounders between houses with lead and other demographics and contribute to the gap and aren't completely controlled — aspects of social class and low mobility that are hard to explicitly capture, it's all old housing. So as the lead level drops and the gap remains steady (increasingly dominated by the confounding factors), more and more IQ gap gets attributed to the small lead-level gap between those living in old housing with abated lead pipes + paint and new pipes + new paint.