Anyone who has glanced at a headline inflation number over the past eighteen months could be forgiven for assuming that the inflation crisis is over. The Consumer Price Index has retreated from its terrifying peak of over nine percent in mid-2022 to a range that hovers in the mid-to-high threes, and the Federal Reserve's preferred gauge, the core Personal Consumption Expenditures index, tells a broadly similar story. Beneath those headline numbers, however, lurks a more complicated and politically consequential reality: the disinflation of the past two years has been overwhelmingly concentrated in goods prices, which have fallen thanks to the normalization of supply chains, declining commodity costs, and the deflationary pressure of Chinese overcapacity. Services prices—the things people buy every day: rent, haircuts, restaurant meals, auto insurance, child care, medical care—have barely budged, and in many categories continue to rise at rates well above the two-percent target that central banks in advanced economies treat as price stability. Understanding why services inflation has proven so stubbornly resistant to the forces that tamed goods inflation is essential to understanding where the economy goes next, and it requires grappling with structural features of the modern service economy that monetary policy alone cannot address.
The Dichotomy Between Goods and Services
The goods-services inflation gap is not a subtle statistical artifact. As of mid-2026, core goods inflation, as measured by the CPI, is running at roughly zero percent year-over-year, meaning that the prices of the tangible products that fill American homes—furniture, appliances, clothing, electronics—are essentially flat on aggregate. Core services inflation, by contrast, is running at around four-and-a-half percent, more than double the Federal Reserve's target and showing remarkably little downward momentum. The divergence reflects a fundamental difference in how prices are determined in the two sectors. Goods prices are largely set in global markets, where international competition, technological improvements, and the constant pressure of lower-cost producers—especially China—exert a persistent downward force. Services prices, with rare exceptions, are set in local markets where international competition is irrelevant, productivity growth is inherently limited, and the dominant cost input is labor, which in a tight labor market rarely falls in nominal terms. A factory in Shenzhen can produce a television cheaper every year; a barber in Chicago cannot cut hair faster without sacrificing quality, and the barber's wages are determined by the local labor market rather than global arbitrage.
The labor-intensity of services is the key to understanding their price dynamics. In many service industries, including health care, education, and hospitality, labor costs account for sixty percent or more of total operating expenses, and labor costs in an economy with unemployment below four percent do not decline voluntarily. Employers may resist wage increases at the margin, but outright wage cuts are extremely rare outside of severe recessions, and in a tight labor market the bargaining power of workers acts as a floor under service-sector wages. Moreover, many service-sector wages are set by multi-year contracts, minimum wage laws, or institutional pay scales that adjust with a lag and embed backward-looking inflation expectations, meaning that even if labor market conditions begin to soften, the passthrough to actual paychecks takes considerable time. The Federal Reserve Bank of Atlanta's sticky-price CPI, which focuses on the components of the index that change infrequently, has remained consistently elevated relative to the flexible-price components, confirming that the inflation problem has migrated from globally-priced goods to locally-priced labor.
The Housing Component That Won't Cooperate
No single category better illustrates the stickiness of services inflation than shelter costs, which account for roughly one-third of the CPI and more than forty percent of the core index. The Bureau of Labor Statistics measures shelter costs using a concept called owners' equivalent rent—essentially, what homeowners would pay to rent their own homes—and that measure has proven to be among the most stubborn components of the entire inflation basket. Despite the cooling of market rents in many metropolitan areas, the official shelter index has remained elevated because the BLS methodology samples rents from existing leases, which adjust slowly as tenants renew, rather than from the marginal rents on new leases, which would capture turning points more quickly. The lag between market rents and the official shelter measure can be anywhere from six to eighteen months, meaning that even as the rental market has softened in real time, the official statistics are still catching up to a reality that is already changing.
"There is a difference between the inflation rate falling and prices falling, and I think the public intuitively understands that distinction even when the economics profession does not always articulate it clearly."
The shelter lag matters enormously for monetary policy because it means the Federal Reserve is essentially flying blind with respect to the single largest component of the inflation index it is targeting. If the central bank waits for the official shelter measure to confirm what the private-sector rent data are already showing, it may keep interest rates higher for longer than necessary, needlessly restraining economic activity. Conversely, if the Fed eases preemptively based on the private data and the shelter measure proves stickier than expected, it risks losing the credibility that it has spent three years rebuilding. Either way, the gap between what the inflation data say and what households actually experience in their monthly rent and mortgage payments is eroding trust in the official statistics and, by extension, in the institutions that rely on them. The practical implications for household budgets also intersect with broader trends in the housing market's new equilibrium, where high financing costs are compounding the affordability crisis.
The persistence of services inflation is more than a statistical curiosity for central bankers to puzzle over during FOMC meetings. It is a direct challenge to the standard macroeconomic models that have guided monetary policy for a generation, models that treat inflation as a single phenomenon that responds predictably to changes in the output gap and interest rates. The reality, as the past three years have demonstrated, is that inflation is not one thing but many, driven by distinct mechanisms in different sectors, and the tools available to central banks are far better suited to managing the demand-side dynamics of goods inflation than the supply-side and structural dynamics of services inflation. As we explored in our discussion of inflation-protected bonds, investors who recognize this divergence have opportunities to position their portfolios accordingly. For households, the practical implication is straightforward: the prices that matter most to daily life are unlikely to come down in nominal terms, and the best-case scenario is that they rise more slowly. That is a message that no politician wants to deliver, but it is the one that the data are telling.
