> the more obvious assumption seems to be that economic output can support more middlemen, so I'm curious what data they used to argue that it works the other w

> the more obvious assumption seems to be that economic output can support more middlemen, so I'm curious what data they used to argue that it works the other way around. That was my conclusion as well - when times are good you don't mind someone who contributes little hanging around. When times are tough, they're out and what little value they offered is replicated by someone desperate to stick around. When a business has a lot of administrative staff times are good and those watching the market can conclude, due to the number of administrative hires, that times are good for a particular market sector. But concluding that it means middlemen are undervalued seems like whoever authored the study had some odd motive.