Confident buyers pay up to 4.4 percent more for homes
A De Nederlandsche Bank working paper finds that consumer confidence strongly predicts entry into the Dutch housing market and raises transaction prices paid by buyers by 2.2 to 4.4 percent in supply-constrained areas. Seller confidence affects the decision to sell but not the price received.
Beliefs open the door and set the price
The authors link individual responses from the Dutch consumer confidence survey (CBS) to 3.3 million housing transactions between 2006 and 2024.
Using a Heckman selection model, they find that respondents reporting negative confidence have 13.9% lower odds of subsequently buying a home, while positive respondents have 12.1% higher odds, relative to neutral respondents.
For sellers, the effects are smaller: a 3.4% lower and 9.7% higher odds respectively.
Conditional on participation, buyers reporting positive confidence pay 2.6% higher prices in supply-constrained markets, falling to 2.2% once selection into the market is corrected for, and rising to 4.4% in the most constrained municipalities.
Seller confidence shows no statistically significant relationship with transaction prices in any specification.
Separating who shows up from what they pay
The paper builds on Bailey et al. (2018) and Kuchler et al. (2023) by directly observing predetermined beliefs from both buyers and sellers rather than inferring them indirectly.
Consumer confidence is measured two to ten quarters before the transaction date to avoid contamination from the decision itself.
The authors exploit cross-municipality variation in developed land shares to separate buyer- from seller-driven price formation, and use marital status as an exclusion restriction in the selection correction.
A narrow window into a broad market
The matched sample covers only 2.3% of buyers and 3.0% of sellers from the underlying survey, raising questions about how representative the price estimates are for the wider housing market.
The reliance on marital status as the sole exclusion restriction is a known weak point of Heckman-style corrections, even if robustness checks support it.
Still, the clean distinction between participation and pricing effects is a genuine methodological contribution to the housing-beliefs literature.