The Economist has an article discussing research on the link between the housing cycle and the investment cycle:

The housing market has generally been both a reliable predictor of downturns and, frequently, a proximate cause. Serious housing troubles preceded nine of the 11 recessions between the end of the second world war and the start of 2020. One exception is the dotcom bust, which was preceded by only a modest housing slump. The other is the recession of 1953, which was triggered by demobilisation after the Korean war. Here housing was a completely innocent bystander.

In the Keynesian model, investment shocks drive the business cycle.  But there’s actually little evidence for that claim, as in almost any business cycle model the investment sector will be highly procyclical.  Milton Friedman‘s “permanent income hypothesis” predicts that people will tend to smooth consumption over time, as their income fluctuates.  Think of a farmer that earns $40,000 in bad years and $100,000 in good years.  The farm family may consume each year as if their income were $70,000/year.

Friedman’s model predicts that consumption will be less cyclical than average, which means that investment will be more cyclical than average.  Even during recessions, consumption spending on food, clothing, haircuts, education, health care, etc., will dip only slightly, even as spending on houses and business equipment falls sharply.

Most recessions (excluding the present slump) are caused by tight money, which reduces NGDP growth.  Because nominal wages are sticky in the short run, employment falls when NGDP falls.  This reduces national output and national income.  Consumption smoothing causes the fall in output to fall disproportionately on the investment sectors (residential and business.)

Housing is almost always an “innocent bystander”, not the actual cause of recessions.