Let’s Get Real About Inflation
There’s a gap between the government’s official numbers and the price levels that Americans actually experience.
By Tom Elliott
Free Expression is a daily newsletter on American life, politics and culture from the Opinion pages of The Wall Street Journal. Sign up and start reading Free Expression today.
Ask any American family whether the cost of living has risen 1.87 times since the year 2000—the official figure produced by the Bureau of Labor Statistics—and you’ll likely get a laugh. That number means a life that cost $100 in 2000 should cost about $187 today. But eggs have gone up by more than that. Same with mortgages. Health-insurance premiums have gone up by a great deal more. Yet 1.87 is the number the Federal Reserve leans on to set interest rates, the number the IRS uses to adjust your tax brackets and the number that determines the annual raise in every retiree’s Social Security check.
The figure isn’t, in the narrow sense, wrong. The consumer-price index is doing exactly what it was designed to do. The trouble is that what it was designed to do isn’t what most Americans assume it does.
The CPI measures the price of a basket of goods after a series of technical adjustments. When something gets more expensive, the formula assumes you switch to a cheaper substitute, and it counts the substitute. When a product improves—a faster phone at the same price—the formula treats the improvement as if it were a price cut, even though your bill didn’t fall. For housing, it doesn’t use home prices at all; it uses an estimate of what homeowners would hypothetically pay to rent their own homes. Each adjustment has a respectable academic rationale. And each one, in practice, has pushed the reported number down rather than up.
Stack those adjustments up, year after year, and they compound into a sizable gap between the inflation the government publishes and the inflation families actually live.
For the past five years I have been trying to measure that gap honestly. The method is simple: Take the government’s own assumptions about how families divide their spending—the very survey the BLS uses to build the CPI—but swap the adjusted prices for the real ones. I take the actual health-insurance premium from the Kaiser Family Foundation’s annual employer survey. The actual cost of shelter—what it takes to buy and finance a home at real mortgage rates, blended with the rent that tenants actually pay—comes from federal housing data, real mortgage rates and market rents. I get real tuition, real driving costs, real grocery and electricity prices from the government’s own retail data. You can find my complete methodology, data sources and chart history at realityindex.co.
The result: That same basket, the one the government says rose 1.87 times since 2000, has actually risen about 2.4 times. In plain terms, the life that the official figure says should cost $187 today really costs around $240—a difference that the headline number never shows you. It’s why the raise that was supposed to keep you even somehow didn’t. It’s the distance between the report on the news and the receipt in your hand.
This isn’t a fringe calculation. I use the government’s own spending weights and the same survey the CPI is built on. The only difference is that I use the real price instead of the adjusted one. I run the same comparison under five different reasonable ways of weighting the basket—including the method the Fed itself prefers—and the gap never closes. It only changes size.
Housing and healthcare account for almost all of it.
The government doesn’t measure what it costs to own a home. It measures what you would pay to rent the house you already own. The reasoning is that a home is partly an investment, and the CPI should capture only the shelter you “consume.” The real cost of keeping a roof overhead—what it takes to buy and carry a home, and the rent that tenants actually pay—has increased for three decades. That hypothetical rent nobody actually pays remains the single largest component in the entire index. The most contested number is also the heaviest one.
Healthcare tells the same story in sharper relief. The government’s medical-inflation figure tracks the list prices of doctor visits and hospital stays. It doesn’t track premiums, which is what most families actually pay. Since 1999, the average family premium for employer coverage has grown to more than 4½ times what it was. During that time, the government’s medical measure rose only about 2.3 times. The premiums families actually pay, in other words, have climbed roughly twice as fast as the official healthcare number admits.
I want to be careful here, because the lazy version of this argument is a conspiracy theory. No one at BLS is cooking the books. The people who build the CPI are serious economists doing careful work, and every methodological choice they’ve made over four decades is defensible. But all of those choices have nudged the reported rate down. Not one has nudged it up. A pattern that points in only one direction for 40 years deserves to be examined in daylight.
The consequences aren’t academic. When the yardstick reads short, every measurement taken with it comes up short, too—by a little each year, and by a great deal over a working life. The Joint Committee on Taxation estimated that the 2017 switch to the slower inflation measure would, by itself, raise federal revenue by $134 billion in its first decade. That’s real money, taken from real households, on the strength of a measurement choice almost no one voted on or noticed.
None of this is meant to abolish the official statistic. The Fed, the IRS and the Treasury will go on using the CPI, and a single official yardstick has its uses. The Reality Index is meant for the family that has suspected, for years, that the number on the news has little to do with the number at the bottom of their receipt—and has half-wondered whether they were the only ones who noticed. They weren’t.
Mr. Elliott is the founder & CEO of Grabien.
Good overview but the article doesn't address the real question as to SHOULD the index act this way? Obviously, a lower number benefits politicians and government but doesn't benefit people directly affected by every growing inflation. So what can be done about it?
Brilliant!