U.S. Inflation Is Normalizing

The temporary acceleration in price increases is already fading. But keep an eye on a few consumer services dependent on low-wage workers.

The temporary inflation spike associated with reopening is already beginning to fade as prices that were depressed during the pandemic continue to normalize and as consumer demand for motor vehicles continues to moderate.

The Consumer Price Index in July was 0.47% higher than in June on a seasonally-adjusted basis. That’s the slowest monthly CPI inflation rate since February 2021 (0.35%) and significantly slower than in June (0.90%). Inflation is currently running just 1.1 standard deviations faster than the January 1995-February 2020 average, compared to 2.7 standard deviations faster in June.


Monthly inflation in July was still faster than the longer-term pre-pandemic average of 0.18%, but, as has been the case since the spring, much of this excess can be explained by temporary factors associated with economic reopening. Those factors have already begun to fade. Here’s the chart I made last month.

And here’s the updated version, with the same scale on the y-axis for perspective.


The most noticeable change is in used vehicles. After prices surged 10.5% in June, contributing about 0.35 percentage points to monthly inflation, prices were essentially flat in July. The surge in prices of used cars and trucks was always likely to stop once consumers were satiated and once the rental companies made progress rebuilding the fleets they had liquidated last spring. We shouldn’t be surprised if prices end up dropping back down to pre-pandemic levels once manufacturers are able to restore production, although that will have to wait until the current shortage of microprocessors ends.

The inflationary contribution of the “reopening categories” was also smaller in July than in June. That’s despite the fact that I included a broader list of categories this month to capture the recent batch of price increases at schools, daycares, live event venues, barbershops, and other sectors that had been hit hard during the pandemic and weren’t yet recovering.

That’s because those gains were offset by the sizable drops in car and truck rental prices and motor vehicle insurance costs in July, which had their biggest one-month price declines since the depths of last spring. Airline fares ticked also lower, while the monthly price inflation at “full service” restaurants (but not “limited service” restaurants)1 and hotels moderated.

So far, this hasn’t yet fed into slower annual inflation. The CPI in July 2021 was 5.4% higher than a year before, just as in June. That’s significantly faster than the 2.3% yearly inflation rate in February 2020.

On the inflationary side, the biggest contributors to the faster yearly inflation rate are oil prices, new and used cars and trucks, most of the sectors hit hard by the pandemic that are now reopening, and a few categories of home goods affected by some combination of higher consumer spending on housing and the chip shortage.

That’s been offset by a few pandemic-affected sectors that haven’t rebounded yet on a yearly basis, plus unusually slow price increases in the cost of healthcare goods and services and in the cost of housing (rent plus “owners’ equivalent rent”, which is an estimate of how much homeowners would have pay to rent their current residence).


Most of these forces should reverse relatively soon—in both directions. As noted above, the monthly data show that most of the pandemic-related categories are already returning to normal.

On the other side, monthly housing cost increases have already returned to their January 2017-February 2020 average after a sharp slowdown earlier in the pandemic. That should eventually show up in the yearly numbers. It’s even possible that rental inflation could accelerate somewhat from current rates to bring rents back in line with their previous trend. (The pandemic-era surge in house prices isn’t a good reason to expect faster rental inflation in the future because that mostly reflected a one-off shift in consumer preferences plus a big drop in mortgage interest rates.)2

Meanwhile, the slowdown in the CPI for medical care is largely due to the way the BLS measures health insurance costs, which is essentially a function of insurer profitability.

Yes, yearly price increases for dental services and hospital services were slightly slower in July 2021 than in February 2020, but that was offset by faster inflation at doctors’ offices and nursing homes. The bigger issue is that the BLS claims that health insurance prices have dropped more than 9% since the peak in August 2020. Before the pandemic, health insurance prices were ostensibly rising more than 20% a year. Expect prices to rise as insurer profits recover and feed through into the BLS model.

While most of the data are consistent with the thesis that the relatively rapid inflation of the past few months is set to reverse as the economy normalizes to pre-pandemic conditions, there are two categories worth following closely that could be a harbinger of something different: fast-food restaurants and domestic services. Prices at both have been rising at yearly average rates of 7-8% since the pandemic began, compared to 1-3% in the year before.

Part of the reason is that employers now have to pay more for labor. Wages for line workers at limited service restaurants have jumped 10% since last summer and have been rising at a 16% annualized rate since the end of last year.

While that’s obviously annoying for owners of restaurants that lack the ability to pass those costs onto customers, it’s unclear what it means for the overall inflation outlook. Notably, prices at sitdown restaurants haven’t risen nearly as much even though wages have grown far faster, likely due to changes in tipping.

I’ll have more inflation analysis later in the month when we get the PCE price data, which is what the Federal Reserve tracks.


The difference between “full service” and “limited service” restaurants is whether you pay before you get your food.


There is a lot more to say on this issue, but I am going to save it for a subsequent subscriber-only piece.