Table 3.

Determinants of price changes

	Price change
	(1)	(2)	(3)
New foreclosures t-1	0.0031${}^{\ast \ast \ast }$	0.0011${}^{\ast \ast \ast }$	0.0005${}^{\ast \ast \ast }$
	(0.0004)	(0.0002)	(0.0002)
New foreclosures t	0.0019${}^{\ast \ast \ast }$	0.0008${}^{\ast \ast \ast }$	0.0005${}^{\ast \ast }$
	(0.0004)	(0.0002)	(0.0002)
New (no forecl.) listings t-1	0.0005${}^{\ast \ast \ast }$	0.0004${}^{\ast \ast \ast }$	0.0003${}^{\ast \ast \ast }$
	(0.0001)	(0.0001)	(0.0001)
New (no forecl.) listings t	0.0001	0.0001	0.0000
	(0.0001)	(0.0001)	(0.0001)
Dummies	Microzone-week	Microzone-week	Microzone-week
Observations	14,765,969	14,765,969	14,765,969
Adjusted R${}^2$	0.0029	0.0029	0.0029

We report only the most relevant parameters (see Table 9 for the other parameters). The dependent variable is an indicator variable taking value equal to 1 if the owner of house i revises the listing price in week t. The number of new nearby listings that are within a radius r of listing i is measured in logs. We consider three different values for r: 150 m (column 1), 300 m (column 2), 450 m (column 3). We control for the type of property, its size, its time on market up to week t and a dummy variable that identify if it is a foreclosure listing. All changes are relative to a baseline probability of adjusting list price of 0.014. ${}^{\ast \ast \ast }$Significant at the 1% level. ${}^{\ast \ast }$Significant at the 5% level. ${}^{\ast }$Significant at the 10% level