Spatial impacts of foreclosure: Do negative spillovers increase with time in foreclosure?

The recent foreclosure surge in the United States started to get worse in 2006 when house prices began to fall as the housing price movement failed to keep in line with economic fundamentals after a prolonged period of house price appreciation. There is a widespread concern that a foreclosure not only drags down the value of the property, but also causes collateral damage to the value of surrounding properties. According to the recent literature, the external costs of foreclosure can stem from various factors such as lack of maintenance, crime, vandalism, and so on. More importantly, the problem of a foreclosure process is not only abandonment, but also uncertain ownership rights. Owners having the prospect of impending foreclosure may lack the time, incentive or money to maintain their homes that they may lose. This in turn may reflect visible signs of neglect, making the neighborhood less desirable to potential buyers.

While the recent literature has treated all nearby foreclosed properties equally in their impact on property values, this paper examines the role that the length of time a property is in foreclosure can play in this regard. This is because the underlying mechanism and the magnitude of such negative spillover impact may depend on the details of the foreclosure process itself. The length of time such a process takes varies from property to property (also from state to state). Consequently, a long duration of the process can have a crucial bearing on the extent of a negative spillover impact as discussed in recent studies. For example, foreclosed properties that are unattended or not properly maintained for a longer period of time are expected to influence surrounding properties more strongly than those in foreclosures for a shorter time period.

This study is based on the housing market in the City of Worcester, Massachusetts. It develops two alternative methods of defining time-variant measures of foreclosure to account for the hypothesized influence of “time in foreclosure” on the strength of such spillover impact. One approach considers those properties that are already foreclosed before the sale of the subject property. The second method chooses those properties for which the filing dates occur before the sale of the subject property, but for which foreclosure occurs after the sale. Both the approaches consider three alternative durations of 0-6, 7-12 and more than 12 months, calculated using the dates on filing and the end of each foreclosure. Foreclosure variables are the count of foreclosures for each time interval. In an effort to examine the research hypotheses, using a variation of a hedonic regression model, this study distinguishes between foreclosures on single- and multi-family dwellings that take place within 660 feet and within 1320 feet of each single-family transaction in the dataset. According to this paper, one would expect the coefficients of these foreclosure variables using both the methods to be negative and their absolute magnitudes to increase with time.