The Real Drivers Behind Rising Prices

The inflation rate is not the problem. Prices rose 3.4 percent over the year ending July 2026, according to the Bureau of Labor Statistics, and core inflation ran 2.5 percent. Those are unremarkable numbers. Households insist things are getting worse anyway, and they are right, because the average conceals which prices moved. The categories a family cannot opt out of have outrun the index for five straight years while wages have gone essentially nowhere. That composition, not the headline, is the story.

The five-year picture is where it shows

Between July 2021 and July 2026, the all-items consumer price index rose 22.3 percent. Shelter rose 27.9 percent. Food rose 25.9 percent, with food at home up 24.2 percent.

Shelter ran 5.6 points hotter than the overall average and groceries 1.9 points hotter. Shelter alone carries the largest weight of any component in the index, and neither category can be deferred. You cannot skip rent for a quarter or wait for a better price on dinner.

Childcare compounds on top. The BLS day care and preschool index rose 3.4 percent over the year to July 2026, matching headline inflation for a category that was already unaffordable.

The wage side is the real indictment

BLS reported real average hourly earnings fell 0.2 percent between July 2025 and July 2026. Over five years the picture is worse in a different way. Nominal average hourly pay rose about 22.7 percent while consumer prices rose 22.3 percent, which nets out to roughly nothing. That five-year comparison is arithmetic on published BLS series rather than a figure BLS publishes.

Workers received five years of raises and bought the same basket. Every household that felt it was running to stand still was reading its own situation accurately.

Housing is doing most of the damage

The National Association of Realtors put the median existing-home sale price at $434,100 in July 2026, up 2.0 percent year over year and the 37th consecutive month of annual gains. The U.S. Census Bureau reported median household income of $83,730 for 2024, the most recent year published.

That puts homes near five times median income. Harvard’s Joint Center for Housing Studies, in its 2026 State of the Nation’s Housing report, states the ratio directly: prices are up 54 percent nationwide since 2020 and remain nearly five times median incomes, against a standard ratio of three that held in the 1990s.

The monthly figure is more vivid than the ratio. Harvard found the payment on a median-priced home reached $3,100 a month in the fourth quarter of 2025, up from $1,700 in early 2020, and that the income required to afford it climbed from $66,000 to more than $120,000.

Renters did not escape. Harvard’s tabulations of Census American Community Survey data show 22.7 million renter households, 49 percent of all renters, spending 30 percent or more of income on housing, and 12.1 million, 26 percent, spending half or more. Cost-burdened renters grew by 2.3 million since 2019, and burdens rose in 44 states.

One statistic in that report does more work than the rest. Renter households earning under $30,000 had a median of $210 a month left after housing, down from $410 in 2019 after adjusting for inflation. By Harvard’s count that describes 13 million renter households, and $210 is the entire budget for food, transport, medicine, and every emergency.

Supply, not sentiment

Housing prices at this level are a quantity problem. Zillow’s analysis of Census data put the national housing deficit at about 4.7 million units in 2024. Other estimates differ substantially because they measure different things. The National Association of Home Builders has put it near 1.5 million counting only vacant units in metro areas, Freddie Mac near 3.8 million including households that never formed, and the National Association of Realtors near 5.5 million based on the construction slowdown since 2000.

The spread is instructive rather than embarrassing. Every method finds a shortage. They disagree about size, not direction.

The affordable end is worse than the total. Harvard’s Joint Center for Housing Studies, tabulating Census American Community Survey data for its 2026 report, found units renting under $1,000 a month in inflation-adjusted terms fell by more than 7 million between 2014 and 2024, while units at $2,000 or more rose from 4.8 million to 10.6 million. The stock did not just get scarce. It moved upmarket.

An honesty check

A piece like this could pick any villain and find a number to fit. So here is one that cuts against the argument: motor vehicle insurance fell 4.5 percent over the year to July 2026. Car insurance was a genuine driver of household cost pain in 2023 and 2024, and commentary that still leans on it is running on stale data.

Energy is similarly awkward. Total energy rose 14.7 percent over the year, with gasoline up 24.6 percent, which is what holds the headline at 3.4 percent while core sits at 2.5 percent. That is a real cost shock, and it is also the category most likely to reverse. Medical care rose only 1.7 percent, with medical commodities actually falling 2.7 percent.

Anyone claiming everything is rising uniformly is not reading the tables.

What the pattern actually is

Strip it down and the mechanism is consistent. Prices in categories with elastic supply and real substitution behave normally. Prices in categories where supply is constrained by physical or regulatory limits, and where households cannot substitute, keep climbing regardless of what the aggregate does.

Housing is supply-constrained by land, zoning, and a decade of underbuilding. Childcare is constrained by staff-to-child ratios that cannot be relaxed without changing the service. Health care is constrained by market structure. Groceries have the least room for substitution of anything a family buys.

Those are the categories running hot. Meanwhile televisions get cheaper, used cars fell 1.9 percent, and the index averages the two together into a number that describes nobody’s experience.

The measurement problem underneath

There is no national living wage figure worth citing, and the reason is instructive. The MIT Living Wage Calculator, which produces the most careful estimates of what a household needs to cover basic costs, explicitly declines to publish a national number, stating the data is meant for highly localized use. Any national MIT living wage figure in circulation is not an MIT publication.

The calculator defines a living wage as what one full-time worker must earn hourly to cover a family’s minimum basic needs where they live, without additional public or private assistance. It counts housing, food, childcare, health care, transportation, internet and mobile, civic engagement, other necessities, and taxes. It excludes savings, leisure, and emergencies.

A subsistence floor, in other words, and one that only makes sense at the county level. That refusal to average is the same lesson the CPI composition teaches. National figures describe a household that does not exist.

The position

Debating whether inflation is 3.4 percent or 2.5 percent misses what changed. Over five years, shelter and food outran the index, real wages gained nothing, the income needed for a median home nearly doubled, and the cheapest 7 million rental units disappeared. None of that shows up in a headline rate, and all of it shows up in a checking account.

Readers tracking these categories together rather than one release at a time can find the longer case for why household costs keep climbing laid out with the same underlying federal data.