Dean Baker argues that an AI investment bust could hit US household spending through a wealth effect that standard economic measures may understate. His explanation turns on housing costs, not just share prices, and it is a provocative case rather than a settled forecast.
Watch Desk analysis
What happened
In an opinion piece for Eurasia Review, Baker starts with a conventional estimate: a 30 per cent fall in the stock market could erase about $24 trillion in wealth and cut annual consumption by roughly $720 billion, assuming households spend three cents less for every dollar of lost wealth. He says economist Jason Furman has argued that the effect in recent years looks closer to one cent on the dollar.
Baker’s proposed explanation is that housing costs have moved out of step with rents. He says the cost of buying a home has risen about 20 per cent more than rents since the pandemic, while national accounts treat home purchases as investment and capture owners’ housing costs indirectly through equivalent rent. Applying that 20 per cent gap to the reported $2.652 trillion value of owners’ equivalent rent would add about $530 billion to measured spending, in his calculation. He suggests this could help explain why the usual wealth-effect estimate appears smaller than recent data imply. Read Baker’s opinion piece.
Why it matters
If Baker’s accounting argument is right, a sharp fall in AI-linked share prices could affect spending through more than investors’ portfolios. He also speculates that falling shares could weaken demand for expensive homes and prompt some owners to cut other spending. That is a possible chain of effects, not evidence that an AI crash is imminent or that a recession would follow.
Our read
This is an interesting challenge to the way economists count housing and wealth, not a crystal ball with a spreadsheet attached. Baker labels the housing explanation speculative, and readers should keep that boundary in view. The useful question is whether the categories capture how households actually experience housing costs, especially when house prices outrun rents.
What to watch
- Whether other economists test Baker’s proposed adjustment against household spending data.
- Whether the gap between house prices and rents persists or narrows.
- Whether AI-related share-price falls begin to affect housing demand or consumer spending.
Discussion spark: Should economists adjust their measures when housing costs move sharply out of line with rents, or would that risk explaining away inconvenient data?
Sources and evidence
- Source update (20 November 2026, 23:59 UTC)
- Source update (20 November 2026, 23:59 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.