That would be Dean Baker, briefly a co-blogger on the old maxspeak, who warns that the recent upbeat reports on housing are not reliable, with them being too much based on highly volatile changes in multi-family housing. For details see http://www.cepr.net/index.php/blogs/beat-the-press/erratic-patterns-in-monthly-housing-starts as well as the closely related http://www.cepr.net/index/blogs/beat-the-press/housing-is-back. Dean must be taken seriously on these matters since his being the first to call the housing bubble all the way back in 2002.
I am not disagreeing with him very much. I agree that for much of the year the hype over a possible double dip in the US has been overdone, which partly explains the sudden surge of enthusiasm we are now seeing at recent high GDP growth rates, which are probably overdone due to being heavily driven by inventory adjustments that are likely to halt after the first of the year, not to mention the continuing likelihood of a European recession with a Chinese slowdown that will put a drag on externally, although those fears have been the main source for all the moaning and groaning in the markets for much of the past few months. And Dean is right on part of the details: single family home construction is barely above its pit in 2009 and not moving; nearly all of this recent increase in housing starts has been for multi-family dwellings, and that is a highly volatile monthly series.
Nevertheless, if one looks at the charts he provides and those he links to, there is a clear upward trend since spring, despite the month-to-month volatility, even if it is mostly in multiple family units. It is also true that it is regional, mostly in the Northeast and West, but any apparently sustained movement should be welcomed. In the chart he shows, with January 2002 as 100, there was a peak in mid-2006 at around 140. The pit in late 2009 was around 20, and it was below 40 this spring. But the November number is at 80. This will probably show declines in coming months, but if the general upward trend continues, this will mean that for the first time since 2006, housing will be a net positive contributor to the US economy, even if only somewhat weakly so.
The newspaper reports have it that rents have been rising in the regions where construction has been rising, reflecting an increase in household formation, with this rising demand finally crashing against the long-depressed supply. Overall housing starts remain far below where they once were, but the issue is direction, and that does appear to be upwards, if erratically and not throughout the US. Dean links to data on rents, but that does not show the most recent that has been reported in the papers.
So, this is at most a mild disagreement with Dean, but in fact I am willing to say that the news on the housing front is improving, at least in terms of a trend for multi-family construction, even if there is a continuing problem in foreclosures and in the single-family home portion of the market. So, while Dean threw in a mention of celebrating Hanukah, I'll throw in one for celebrating Christmas as well, :-).