Felix Salmon, looking at the various Case-Shiller home-price indices, puts it pretty well, I think:
If housing kept track with CPI inflation, the Case-Shiller index would be at 125 now; in fact, it’s at 140. But of the 20 cities on the Case-Shiller list, just 9 have managed to outperform inflation: Boston, Los Angeles, Miami, New York, Portland, San Diego, Seattle, Tampa, and Washington. The big outperformers — New York and Washington — more than make up for the underperformers like Detroit, Cleveland, and Atlanta.
My gut feeling is that this means New York and Washington have significantly further to fall, in terms of housing prices; even Miami, at 144, is still looking pretty rich. San Francisco might look cheapish at 120, but it was artificially inflated, at the beginning of 2000, by the dot-com bubble: just a year earlier it was at 85.
Overall, I think people looking for a bottom here are being premature. There’s still a huge overhang of unsold housing, and it’s still very hard to buy a house, if you don’t have a large down-payment — and given the US savings rate over the past few years, not so many people have that sort of money to hand. The precipitous part of the decline might well have come to an end: from here on in we might see a slow grind lower over many years. Only if you can live with that kind of long-term price decline should you be even thinking about buying a place right now.
For what it's worth, the latest Case-Shiller release notes that the broad housing market index is about where it was back in 2003. That suggests to me, as it does to Salmon, that we've still got a little ways to go, adjustment-wise, and that it's probably a bit premature for the monthly price-increase the index is reporting.