As a programmer in my 50s, this is of concern to me.

But I think ProPublica is putting their thumb on the scales a bit. They add in fudge-factors for people who say they are retired, but ProPublica suspects were pushed out. They further add in some skew for those who say conditions deteriorated, but ProPublica assumes were really shown the door. This is just adding weight to their natural bias-- it's no way to actually learn anything.

P.S. I lived for 19 years in South Dakota. I cannot image a basement that is overheated in December.