Discover more from Discourse
The Government Isn’t Cooking the Books on Inflation
By Christopher M. Russo
The Consumer Price Index rose by 0.9% in June, or over 11% at an annual rate. For many of us used to decades of consistently low inflation that sounds like a lot, and it is.
While inflation is raging, talk about inflation is all the rage, with increasing concern that the U.S. may be entering a period of higher inflation. But what exactly is inflation? How is it measured? And are those measurements accurate?
The following Q&A answers these and other questions about the economic phenomenon economist Milton Friedman once called “taxation without representation.”
What is consumer price inflation?
Imagine you’re in the grocery store and you fill your cart until you’ve spent everything in your budget. Next time you shop, you notice that the quality and prices of the groceries have changed. “Groceries inflation” measures how much extra money you’d need to spend to be indifferent to these changes. For example, if you need 10% more money in your budget to compensate for price increases or quality decreases, then groceries inflation is 10%.
Of course, we consume more than just groceries. Sometimes we eat out! We also consume energy like electricity, gas and fuel; durable goods like clothing, vehicles and housing; services like medical care and education; and much more. Change the thought experiment to a superstore with absolutely everything that we consume. That’s “consumer price inflation.”
Can you illustrate this definition?
Suppose current goods and services (which, for brevity’s sake, I’ll refer to simply as “goods” from here on out) are identical to prior goods. If all prices doubled, meaning that they rose by 100%, then inflation would be 100%. But it would be imprecise to reason that inflation itself is “a general rise in prices.” To show why, suppose prices rose by 10% on average. Some grew faster, while others grew slower or even fell. People would comparison shop to find substitutes, like buying the store brand instead of the name brand. This “substitution effect” lowers inflation.
Suppose instead that current prices are identical to prior prices. If all goods shrink by half, meaning that they’re 50% less, then you’d need twice as much money to buy the same amount. That means that inflation is 100%. The substitution effect applies in this case as well. If goods only shrink by 50% on average, then people would still comparison shop. Notice, too, that there’s more to quality than just size. The quality of a box of cereal might depend on the cereal’s taste, the attractiveness of the box, and anything else consumers desire.
In practice, across all the goods we consume, price and quality are always changing. Inflation captures both through the lens of consumer choice.
How do governments measure inflation?
Since the “superstore” with all goods and services doesn’t actually exist, governments construct proxies to measure inflation. In the United States, two popular ones are the Consumer Price Index (CPI) and the Personal Consumption Expenditures (PCE) price index.
At a high level, the calculations involve applying economics and statistics to data on the prices and quantities of goods bought. Each step requires assumptions and choices. For example, do you survey households for data on their purchases, or use a sample of transactions? How do you handle consumable goods that are purchased infrequently (e.g. housing) or those without market prices (e.g. most medical services)? How do you adjust for differences in quality? How do you smooth over predictable seasonal variations, if at all?
Different answers lead to different measurements, each imperfect. For example, CPI inflation may underestimate substitution effects because its basket is based on an annual survey of consumers. PCE inflation more readily reflects consumer substitution because it is based on monthly personal consumption. On the other hand, both CPI and PCE may underestimate inflation because of the difficulty of measuring the market price of owner-occupied housing.
From an economics perspective, homeowners pay an opportunity cost for not renting out their home. As their home value appreciates, so does their opportunity cost, since if the home were rented, owners could charge more. That raises inflation as much as actual rent increases paid to landlords. It’s another case of “no such thing as a free lunch.”
How does inflation relate to changes in the money supply?
The price level is proportional to the money supply. If economic output and people’s demand for money remain fixed, a doubling of the money supply will double the price level. In other words, more money will be chasing the same goods. Even so, output and money demand can move alongside the money supply. Even in Friedman’s famous statement that “Inflation is always and everywhere a monetary phenomenon,” he continued on to explain, “in the sense that it is and can be produced only by a more rapid increase in the quantity of money than in output.”
By one common measure, the money supply grew by 32% from February 2020 to May 2021. Over the same period, CPI inflation was 3.8% and PCE inflation was 3.3%, roughly 10 times less than the rate of money growth.
Couldn’t the government just be cooking the books to hide double-digit inflation?
All governments—including our own—are run by men not angels. In order to hide its out of control inflation, Argentina cooked its consumer price index from 2007 to 2015. In a 2012 paper, MIT economist Alberto Cavallo constructed an alternative consumer price index. By his measurements, the inflation rate in Argentina was nearly three times higher than the official rate.
But for the United States, it’s not plausible that cooked books hide double-digit inflation. Here’s some quick math. Consider U.S. GDP measured in dollars. The Bureau of Economic Analysis measures that dollar GDP grew by about 2.3% from March 2020 to March 2021. Over the same period, the money supply grew by 24%. If the price level is also up by 24%, then real GDP must be down by about 18%. That’s absurd and clearly not true: We’d need Depression-era soup kitchens to feed all the unemployed.
Even so, inflation has surged in recent months. Is inflation about to take off?
Near-term inflation forecasting is notoriously hard, even for the Fed. These forecasts are often based on statistical patterns, meaning that we’re just making a best guess based on recent inflation and other data such as energy prices. The Cleveland Fed’s Inflation Nowcast is one such attempt.
As of July 9, the Cleveland Fed’s model foresees inflation slowing to the Fed’s 2% target in June and July. However, the model experienced a large forecast miss: June’s CPI was over 11% at an annual rate, compared to a forecast of 4.4%. “Annual rate” expresses how much inflation there would be if that month’s rate carried over an entire 12-month period. In the same way, if you drive 5 miles in 5 minutes, then your average speed is 60 miles an hour.
That said, the Federal Reserve doesn’t attempt to tweak inflation month by month. It can’t fix supply chain disruptions or pandemics. Rather, the Fed aims to achieve 2% PCE inflation in the long run. Economic theory suggests that the Fed can accomplish this objective if it’s committed to it. If inflation rose above 2% on a sustained basis, then this commitment would require policymakers to raise interest rates.
However, if the Fed prematurely raises rates in response to temporary supply-side pressures, then it risks kicking the economy back into recession. With interest rates already close to zero, it is harder for the Fed to combat recessions than inflation. There’s just less space to lower interest rates. For that reason, the Fed has remained patient on inflation so far.