Fig. [ ]. Source: NBER Macrohistory Database
The pattern of private spending on residential construction as opposed to nonresidential suggests that its decline that started in 1927 and accelerated in 1929 may have largely stemmed from the satiation of the housing market with new housing units rather than the economic downturn of 1929 because of the sheer magnitude of the increase in residential construction spending between 1920 and 1926, especially compared to the rise in its nonresidential counterpart.
The evidence that the accelerated decline in residential construction spending started in late 1928 is twofold. First, as we have just shown, residential construction spending had already been decreasing since 1927. Secondly, although there are no monthly data on spending on private construction components, we can try to infer conclusions from the monthly data on total construction value for 37 states published by NBER.