This article provides a "teachable moment" for urban economists. Consider a low home price neighborhood where one home owner defaults on his loan and gives the keys to Bank of America. BOA faces a transaction cost issue that it doesn't have anybody on payroll who can at relatively low cost go to the property and check whether this asset is "healthy" over time. For example, if hoodlums grow pot there or if there is a fire --- how would BOA know this?
Does BOA have an incentive to invest money in maintenance and upkeep? Suppose that it would cost $100 a month to mow the grass and maintain the home. If the home's value is expected to decline by 2% a year (because of the ongoing recession) and the home is worth $250,000 then BOA expects to lose $5,000 a year in asset value. If BOA doesn't view the $1,200 in maintenance expenses as leading to a higher rate of return then it won't make this investment and the home's quality will decline. Now, the spatial externality here occurs if the homes next to the depreciating home that BOA owns decline more in value because they are next to a "nasty home". Economists have been estimating hedonic pricing models and augmenting the statistical model to include a count of homes in foreclosure in a vicinity of a given home. Such statistical models do not attempt to pinpoint why foreclosure has a causal effect on nearby home values. It is possible that a 3rd variable (a neighborhood shock) causes both events rather than the causality being that local foreclosure spills over and lowers nearby home values but this story is intuitive for homes very close to the "pot den". If I live 7 homes away from a foreclosed house, I don't really see how this can lower my home's price unless you believe in a "domino effect" story and my home is the 7th domino.