A very interesting debate in the comments section of the Northern Virginia Housing Bubble Fallout Blog over whether Arlington prices, which have held up reasonably well post-bubble, are likely to fall due to the "substitution effect". With housing prices having plummeted in the outer suburbs, will home buyers, for example, decide to buy a 200K home in Prince William County instead of a 700K home in close-in Arlington?

My best guess, despite being an Arlington homeowner, is that Arlington is due for a fall. Arlington has many advantages, and the disproportionate rise in Arlington prices has some legitimate reasons, including how much nicer and safer DC has become, reasonably wise development policies in Arlington, and the general trend toward Yuppies wanting a more urban experience. But Arlington also has uneven schools, older housing stock, small lots, very high prices, and nasty traffic problems for those who need to commute west (like my brother-in-law, who spent an hour each way driving from Courthouse to Fairfax City each day, a fifteen to twenty-minute drive with no traffic), where the high-tech companies are located. At some point, marginal buyers will prefer neighborhoods in the outer burbs where prices are at 2002 levels (down 50% or so from the peak) to neighborhoods in Arlington where prices are still at 2004 levels (down 10-15% from the peak).

UPDATE: A major reason prices have fallen so hard in the outer 'burbs, but not Arlington, is the much great incidence of foreclosures in the former. Foreclosures, and for that matter short sales, have dragged down the prices of "normal" sales, creating what may turn out to be short-lived bargains. If I were buying for investment now, it strikes me that an 80K townhouse in Manassas or Dumfries that could be rented for $1,200 a month is a much better deal than an $800K house in Arlington that rents for $3,200.