2024-12-01-美联储-锁定_加息_住房市场和流动性（英）_59页_2mb

Analysis: "Locked In: Rate Hikes, Housing Markets, and Mobility"

Key Findings

• Lock-In Effect on Mobility: Rising interest rates in 2022 widened the gap between existing mortgage rates and market rates for many homeowners. This discouraged moves, accounting for approximately 44% of the observed decline in mobility between 2021 and 2022, primarily affecting homeowners "out of the money."

• Nature of Reduced Mobility: The reduction in mobility is predominantly concentrated on short-distance moves within the same metropolitan area (CBSA) or within a 30-mile radius. These moves appear discretionary, less job-related, and often tied to consumption (e.g., moving up the housing ladder, changing household structures). Long-distance moves, typically job-related, were largely unaffected by rate gaps.

• Housing Market Impacts: Exposure to the lock-in effect led to:

  • Declines in Listings: Fewer homes listed for sale in markets more affected by lock-in, likely due to reduced churn.
  • Increased Prices: Higher house price growth, particularly in tighter (already seller-favored) housing markets. The mechanism involves reduced supply meeting demand in already constrained markets, driving up prices.

• Labor Market Consequences: Contrary to concerns, the lock-in effect did not significantly restrict workers' ability to move for jobs or cause substantial labor misallocation. The decline in mobility was largely driven by consumption-related moves rather than job-seeking.

Mechanism

A housing search model illustrates that lock-in reduces within-market churn. In tight markets (more buyers than sellers), reduced churn increases market tightness (raises the buyer-to-seller ratio), which drives up prices. Existing data confirm stronger price and listing effects in initially tighter markets following lock-in.

Sample & Data

The study utilized a novel, large-scale dataset merging loan-level mortgage data (Equifax CRISM), property deeds (CoreLogic), and market data (ICE McDash, Realtor.com, Freddie Mac PMMS) from 2009 to 2023, covering 2.5 million mortgages.

Note: Standard errors and definitions were controlled for various borrower characteristics, loan terms, borrower types, and time effects.