That’s what Ed Leamer noted in 2007. Residential construction employment, spending, housing starts, new home sales, all normalized to 2021M11.
Figure 1: Residential construction employment (blue), residential construction spending deflated by PCE (tan), housing starts (green), new sales of single-family dwellings (red), all seasonally adjusted, in logs 2021M11=0. Source: BLS, Census, Census/HUD, BEA and author’s calculations.
back in September, Leaguer did not see an impending recession. I don’t know what we would do with the current data. In fact, he argues that housing is a less reliable predictor of recession, given the lack of strong housing accumulation in the face of continued strong demand, and the general strength of the financial system, unlike in 2007.