The most-cited number in U.S. housing economics is the price-to-income ratio — the median sale price of a home divided by the median household income. It is a useful number because it is simple, comparable across decades, and doesn't require you to choose which interest-rate regime to anchor against. It is also a misleading number when read in isolation, because the ratio that determined whether a buyer could actually qualify for and carry a mortgage looked nothing like the ratio printed in the textbooks. The full affordability story is the price-to-income ratio and the mortgage rates and the down-payment burden, evaluated together.
This explainer walks through six decades of that joint calculation, using the underlying U.S. data — Census new-home medians, NAR existing-home medians, Census household incomes, and Freddie Mac PMMS mortgage rates — to show what each decade actually looked like for a household at the middle of the income distribution buying the median home.
What price-to-income gets right, and what it misses
Price-to-income captures the absolute scale of a housing transaction relative to household earnings, and it strips out interest-rate noise. A 4.0× ratio in 1981 and a 4.0× ratio in 2021 both mean a buyer at the median is spending four years of pretax income to buy a median home. But the rate at which they are financing the purchase, and therefore the share of monthly income going to housing, can differ by a factor of two or more. The 1981 buyer at 4.0× would have been paying nearly half of gross income to principal-and-interest; the 2021 buyer at 4.0× would have been paying about a fifth.
This is why affordability indexes that rely solely on price-to-income systematically misjudge the 2020s. By that single metric, 2024 is comfortably the worst affordability environment in the modern record. By the joint metric — payment-as-share-of-income — 2024 is meaningfully worse than 2005 and 2021 but materially better than 1981. The economic reality lies somewhere between those two readings, and the rate at which the rate-lock era resolves will determine which one the late 2020s converge toward.
The 1960s and 1970s: the normal range
Through the late 1960s and early 1970s, the U.S. price-to-income ratio sat in a remarkably stable band of roughly 2.5× to 3.0×. The 1968 existing-home median was $20,100 against a $7,743 median household income — a ratio of 2.60. The 1973 reading was 2.75; the 1971 reading, with the first Freddie Mac PMMS data, was about 2.89. At the prevailing 7–8% mortgage rates of the period, monthly P&I on an 80% LTV mortgage ran 15–18% of gross income for a household at the median.
That arithmetic produced the highest U.S. homeownership rate up to that time and an environment in which a single-earner household could plausibly buy the median home with a one-year emergency fund and a normal employment record. It is the baseline against which every subsequent decade's affordability measures itself.
The mid-to-late 1970s started moving the band. By 1978 the existing-home median had reached $48,700 against a $15,064 income — a ratio of 3.23. Mortgage rates had risen to 9.64%. Monthly P&I as a share of income climbed to roughly 25%. The cycle of rising prices outpacing income gains had begun, and the ratio that would persist into the 1980s started locking in.
The 1980s: when the rate dominated everything
The Volcker era is the clearest demonstration in the U.S. record that affordability is not a function of price-to-income alone. The 1981 existing-home median of $66,400 against a $19,074 household income produced a price-to-income ratio of 3.48 — only modestly higher than the 1978 reading. But at the 1981 average mortgage rate of 16.63%, monthly P&I on an 80% LTV mortgage was roughly $742, against monthly gross income of about $1,590. Forty-seven percent of gross income to housing principal-and-interest alone, before taxes, insurance, or maintenance, is a level at which prudent underwriting cannot occur and existing-home transaction volume collapses. It did: existing home sales fell from 3.99M (1978) to 1.99M (1982), the deepest peak-to-trough decline in the modern record.
By the mid-1980s, mortgage rates had begun to fall but prices had not, so the ratio stayed elevated even as payment-burden eased. The 1985 reading: 3.20× price-to-income, ~37% payment-to-income at 12.43% mortgage rates. The decade closed with the 1990 reading at 3.07× — the lowest of the decade — but with payment burden still at ~30% even as rates fell into single digits. The Volcker housing crash explainer walks through how the asymmetric path — prices stuck while rates moved — produced the era's distinctive affordability profile.
The 1990s: rates fall, ratios stabilize
The 1990s were affordability's most underappreciated decade. The price-to-income ratio held nearly flat — 3.07× in 1990, 3.24× in 1995, 3.31× in 2000 — even as nominal home prices rose from $92,000 to $139,000. Income growth tracked the price growth almost step for step. Meanwhile mortgage rates fell from the 10.13% 1990 reading to a 6.94% 1998 trough, the lowest rate environment in 25 years. Monthly P&I as a share of income fell from ~30% to ~22% over the decade.
The combination — stable ratio plus falling rates — produced the longest sustained homeownership-rate expansion of the post-war era, from 64.0% in 1990 to 67.5% in 2000. Refinancing volume, household-formation rates, and first-time-buyer participation all reached cycle highs. The 1990s housing recovery explainer documents the post-S&L cleanup and the demographic expansion that drove this period.
The 2000s: the boom that reset the baseline
The 2000s broke the post-1971 stable-ratio pattern. From a 3.31× reading in 2000, the price-to-income ratio rose to 4.73× in 2005 at the cycle peak — the highest reading the U.S. had recorded since the data series begins. Mortgage rates were unusually low for the period (5.87% in 2005), so payment burden remained manageable on the surface — roughly 27% of gross income at the 2005 peak. The misleading aspect of the 2005 calculation, only visible in retrospect, was that the underwriting process supplying those mortgages was incorporating zero-down structures, interest-only periods, and stated-income documentation. Households at and above the median appeared to qualify for purchases at 4.7× their income only because the verification of that income had been quietly decoupled from the mortgage product.
The 2008–2011 unwind reset prices but not incomes. By 2011 the existing-home median had fallen 24% to $166,200 while income held roughly flat at $50,054. Price-to-income returned to 3.32× — the lowest reading since 1990. Rates fell to 4.45%, the lowest in the PMMS series at that time. Payment-to-income dropped to roughly 21%. From any pure-affordability standpoint the 2011–2012 window was one of the best entry points in the post-war era. The 2008 subprime collapse explainer covers how an ostensibly affordable market was unaffordable on credit-availability grounds — a recurring theme.
The 2010s: incomes climb, prices climb faster
Through the 2010s the price-to-income ratio drifted from 3.32× (2011) to 3.96× (2019). Monthly payment burden held in the 22–25% range as mortgage rates stayed unusually low (3.65% in 2016 — the lowest annual reading until 2020 broke it). Prices outpaced income growth by roughly 1.5 percentage points annually for most of the decade, but the rate decline absorbed the gap. The decade's affordability arithmetic was a slow leak rather than a sudden shock.
By 2019 the joint metric — 3.96× ratio, 24% payment burden at 3.94% rates — was modestly worse than 1971's baseline but historically normal. The bottom 30% of the income distribution had begun to be priced out of homeownership in coastal metros, but the median U.S. buyer remained well within the affordability envelope. The setup for the 2020s shock was already visible: any sustained move higher in rates with the existing price level intact would push payment-to-income to 30%+.
The 2020s: the compound shock
The 2020–2024 sequence was the steepest deterioration in measured affordability in the modern data record. The trough-to-peak run was driven by two compounding moves: the median existing-home price rose from $295,300 (2020) to $408,000 (2024), a 38% nominal increase in four years; the 30-year fixed mortgage rate rose from 2.96% (2021) to 6.84% (2024), a 388 bp move. Income growth could not absorb either piece in isolation, let alone both.
The arithmetic at each end of the window: a 2021 buyer at the median financed a $285,680 loan at 2.96% for a $1,196 monthly P&I — about 20% of $70,784 gross income. A 2024 buyer at the median financed a $326,400 loan at 6.84% for a $2,135 P&I — about 31% of $83,730 gross income. The same household-income percentile, attempting the same purchase decision four years apart, faced a payment that consumed half-again as much of monthly cash flow.
The 4.87× price-to-income ratio in 2024 was the highest reading since the modern data begins. The 31% payment-to-income reading was higher than 2005's peak but well below 1981's 47%. The middle-decile household in 2024 faced a worse housing affordability environment than at any point in the post-2005 era, but a meaningfully better one than at the Volcker peak. The rate-lock era explainer walks through how this affordability profile interacts with the seller-side constraint that has reshaped supply.
How the 2020s normalize: three rough paths
The compound-shock environment can resolve along three paths, each measured in years rather than months:
- Rates ease. A move from 6.84% back to roughly 5.0% — broadly the 2018 environment — would cut monthly P&I on an unchanged $326,400 balance by about $385. That alone would drop the payment-to-income ratio from 31% to about 25%, a meaningful normalization without any price or income change. A move below 4.5% would dissolve the affordability problem at current prices.
- Incomes catch up. Median household income at 4% nominal annual growth for five years would put the 2029 reading near $102,000 against a flat $408,000 price — bringing the price-to-income ratio to about 4.0× and the payment-to-income ratio at unchanged rates to roughly 25%. This is the slowest of the three pathways but is the one that does not require any policy change.
- Prices give back. A nominal price decline of 10–15% — comparable to the 2008–2011 unwind in scale — paired with stable rates and modest income growth would close most of the gap by the late 2020s. This is the most economically painful pathway, falls disproportionately on recent buyers carrying high-LTV mortgages, and historically requires either a recession or an inventory shock to occur.
The actual path will be a mixture of all three. The data series — six decades of joint affordability readings — suggests the ratio reverts toward its long-run average over five-to-seven-year windows even when the deterioration is severe. The 1981 ratio took until 1990 to fully normalize; the 2005 ratio took until 2011. The 2024 deterioration is younger than both of those precedents and earlier in its mean-reversion path.
Price-to-income alone tells you the rough scale of the problem. Payment-to-income tells you whether the median household can carry the median mortgage. Both readings have to be read together, and the 2020s are the first decade where they have diverged this far.
For prospective buyers, the practical takeaway is unambiguous: the affordability question in the 2020s is overwhelmingly a rate question, not a price question. The same physical home at unchanged price becomes 25% cheaper to carry per month if rates fall 200 basis points. The purchase calendar that matters is the rate calendar.