A Little Housing Mystery

Home prices in the US right now look like they still have a ways to fall, according to a graph that's been making the blog rounds lately:

Bryan Caplan points out that this is much worse than the experience in the Great Depression:

Home prices did amazingly well during the Great Depression. According to Schiller's index, it looks likes inflation-adjusted prices fell from about 74 to 69 between 1929 and 1933 - about a 7% decline. By 1940, they were up to about 82. The double-dip recession of 1937-8 shows up as a small downward blip in the housing market, nothing more.

What gives? I realize that nominal housing prices must have declined massively during the rapid 1929-33 deflation. But the resilience of the housing market in the depths of the Depression is still most puzzling.

Part of this is that house prices simply didn't have so far to fall, because they hadn't run up so sharply.  But still, it's mysterious, because other measures--like employment, GDP, and industrial production--had fallen much more sharply.  Why didn't houses suffer so much?

I can tell a bunch of stories--the housing market was much less dependent on debt at the time, and people were less likely to live alone, so they weren't so vulnerable to a single income shock.  But I don't have any proof, so they're just stories.