What Nominal Mortgage Rates Can—and Cannot—Tell Us About Affordability

The subject is residential mortgage affordability under changing nominal interest rates, and the final judgment is more precise than either “only real rates matter” or “nominal rates explain everything.” A fixed, multi-period mortgage can make a higher nominal rate reduce the principal a new borrower can carry, even when the real rate is unchanged, because a current payment-to-income constraint turns the nominal payment into a borrowing limit. The evidence supports that mechanism across several forms of analysis, while also showing why the effect cannot be treated as a universal estimate for every borrower, loan, city, or economic outcome.

The central answer is a contract-and-constraint answer

A nominal rate can tighten affordability on its own in a conditional but economically meaningful sense. “On its own” means that, after holding the real-rate comparison fixed, the nominal rate can still change the required dollar payment on a new multi-period mortgage. It does not mean that nominal rates move independently of inflation, monetary policy, income, house prices, or expectations in actual economic history. Those forces often move together. The claim is that the mortgage contract creates an additional channel within that entangled environment.

The channel begins with the payment schedule. A fixed-rate mortgage requires a sequence of nominal payments over many periods. For a given principal and term, a higher quoted rate requires a larger payment. If a lender or borrower limits the share of current income devoted to that payment and to other obligations, the larger payment reduces the principal that can be supported. If the constraint binds, the household must borrow less, buy a less expensive home, provide more funds, reduce other debt, or leave the transaction for later. The real rate remains relevant to the cost of shifting purchasing power through time, but it is no longer the only rate that can affect the feasible purchase.

The paper’s stylized example shows the logic without pretending to describe a typical mortgage. A household earning $100,000 is assumed to face a 50 percent first-year payment ceiling, or $50,000 annually, after other debt payments and property taxes are considered. At a zero nominal rate and zero expected inflation, that payment supports roughly $1.5 million over a 30-year term. If the real rate remains zero but both expected inflation and the nominal rate are 10 percent, the same ceiling supports about $470,000. The figures are hypothetical; the comparison is the evidence-bearing part. A fixed current-dollar ceiling can support a much smaller principal even when the real-rate condition is held constant.

That example also fixes the scope of the conclusion. The nominal rate is most likely to be the active margin when the payment constraint binds or nearly binds. A household with substantial room below the lender’s limit may face a higher cost without changing the approved principal. A household whose desired purchase is already limited by the real rate may not be affected by the additional nominal channel in the same way. The paper’s conclusion is therefore about a contract interacting with a constraint, not about a quoted rate mechanically determining all housing demand.

The empirical evidence is strongest when read as triangulation rather than as one oversized estimate. The Michigan Survey of Consumers asks whether respondents think it is a good time to buy a home. In a baseline regression covering 228,425 observations from February 1979 through June 2025, a one-percentage-point increase in the national 30-year mortgage rate is associated with an approximately 11-percentage-point lower probability of a favorable answer, represented by a nominal-rate coefficient of about -0.11. The respondent-specific real rate is the nominal rate minus expected inflation over the next five to ten years. Conditional on the nominal rate, its coefficient is positive in the reported specification, an unusual sign that should not be converted into a general claim that higher real rates improve housing demand.

The survey result speaks first to what households report. It is compatible with people noticing the nominal payment conditions of a new mortgage, but it is not a mortgage-demand elasticity or an approval rate. The paper adds unemployment and expected house-price inflation controls, and the nominal-rate coefficient changes little.