Interesting quote from the original article
Debt in collections originates from non-payment of a bill, and includes events such as failing to make payments on an outstanding credit card balance, not paying medical or utility bills, or even failing to pay a parking ticket. Importantly, a debt that is reported as in collections can remain on a person's credit report until the debt is seven years past due. Unlike debt past due, debt identified as in collections in our data has not necessarily been verified in the past 12 months.
So if you fail to pay a parking ticket, you are part of this list for seven years? I think this really distorts this data. If I were the analyst, I would run a separate report that both cuts the time down and eliminates or significantly discounts small debts. When I think of "debt in collections", I'm thinking thousands in unpaid bills, not a forgotten parking ticket.
The more I read the original paper, the less I think of the conclusion although I don't doubt the math. From the paper, the process is debt goes past due, then after 180 days into collection. The percentage past due is around 5% of people, but somehow we get to 35% in collections? Shouldn't we have a lot more past due than in collection? The data also doesn't include things like pawn shops where you might expect a high default rate. It also doesn't make sense that Americans are getting their credit card spending under control (an easy source of unsecured debt) but are seeing a high default rate. Wouldn't someone borrow on their credit cards to make the bills if they had to? Something just doesn't close here.