I wish I could remember the details, but one of the articles I read about this went a step further and suggested when the sell side is limited (a single company

I wish I could remember the details, but one of the articles I read about this went a step further and suggested when the sell side is limited (a single company moving to a region like Amazon) and the benefits are hard to estimate (community utility) then the best bid will inevitably be deriving negative utility due to over-bidding. In your example, one city might bid 11 because they wanted to beat 9 and estimated the total utility at 13, not knowing it would be 10.