Round numbers may be bad for business when it comes to selling a house, according to three profs from Cornell University. The current issue of The Atlantic Monthly picks up the story:

The authors showed their subjects a listing for a house along with various prices, and asked whether those prices seemed high or low. Precise prices like $391,534 were seen as cheaper than round ones like $390,000, even though the round prices were actually lower. The authors then examined more than 27,000 real-estate transactions on Long Island and in South Florida and discovered the same effect at work in real-life deals. In South Florida, having at least one zero at the end of the list price lowered the final sale price by about 0.72 percent compared with houses listed at a similar price, three zeros lowered it by 0.73 percent, and each additional zero lowered it another 0.39 percent.

Can any readers back this up with anecdotal evidence?
Primary Sources: Pound Foolish [The Atlantic]