A bad joke about economists goes something like this: if you ask a lawyer what two plus two is, he will say four. If you ask a doctor, he will also say four. But if you ask an economist, he will stroke his chin, make some calculations, and reply that the answer lies somewhere between three and five.

To be fair to economists, the explosion in the number of ways we measure the economy allows for a wide set of views about the same phenomena. Inflation during and after the pandemic is a good example.

Some looked at the latest Consumer Price Index (CPI) release last week and saw easing price pressures (the headline index declined from a year earlier). Others saw the opposite (the core index increased from a month earlier). Today, an economist being asked the two plus two question might likely reply: it depends, what measure of two are you using?

Gauging inflation shouldn’t be complicated. Historically, and for the most part, a single number has captured how quickly prices are rising: the annual change in the CPI, as measured by the Bureau of Labor Statistics. 

That said, even in normal times, economists have made a plethora of adjustments to that single piece of data to make it more reflective of the true face of price changes. They have measured changes over a shorter horizon (one or three months, for example). They’ve made adjustments for seasonality—and removed volatile components, like food and energy—to smooth out the trend (arriving at a “core” measure of inflation). More recently, some gauges have begun stripping out monthly outliers to reveal a “trimmed mean” of price rises.

But the pandemic has introduced the need for further adjustments to account for a burgeoning set of idiosyncratic, pandemic-bloated price jumps.

Snarled supply chains reduced the availability, and boosted the price, of new cars. As we retreated from subways and moved to rural areas, used car prices soared. And then, as the economy opened and the government mailed out stimulus checks, airfares soared.

These costs are all important—after all, people are genuinely paying for them. But they can also obscure the underlying forces influencing inflation on a more permanent basis—the ones driving policy and financial markets. To this end, economists have sought to remove these pandemic-influenced components from inflation measures.

But accuracy can come at the cost of complexity. Accounting for the different constructions of the index, the various possible adjustments, and the time horizons along which changes in prices can be measured, there are now thousands of ways to answer the same questions: how rapidly are prices rising? Which components are driving prices higher? Is the trend accelerating, steady, decelerating, or inflecting?

Visualizing the data helps. The chart below shows the rate at which prices are increasing for the major spending items for urban consumers, compared to 2019. The bubbles are sized according to each items’ weight in the consumer price index. They are colored according to whether they rose at a faster monthly rate than in the prior month (yellow), or a slower one (blue).

At first glance, the data is not reassuring—most bubbles are above the x-axis: these items continue to rise in price at a faster rate than in 2019. And most are yellow: the pace at which they are climbing accelerated in the last month (the number of items accelerating is higher than normal). And some of the large categories, like measures of shelter costs—which include rent, and owner-equivalent rent (the imputed cost of owning a home based on rental prices)—are both higher than normal and accelerated over the last month. 

But markets are the best arbiter of whether a piece of economic data is good or bad news. In this CPI report, investors liked what they saw. Following the release, the S&P 500 jumped almost 1%. Yields on government bonds dropped, indicating that investors expect cooler inflation to allow the Fed to bring an end to rate hikes.

What explains investors’ upbeat reaction?

For one, some of the bad news in the CPI report came from items that have lost their sting. In particular, the government’s measures of shelter costs in CPI have been superseded by private measures. Zillow, the online real estate listing company, provides monthly price indices for new rentals. These lead CPI shelter inflation by around half a year. Zillow’s indices have been declining since early 2022. Investors already know where shelter prices are heading.

Aside from shelter, there are two other large components of core CPI. The first is goods inflation—the rate of price increases for items such as cars, clothes, and medicines. These are rising at about a 2% annual rate. Goods are not a big drag on overall inflation, as they were in the two decades prior to covid, but neither are they boosting it above the Fed’s target.

The second large component of core CPI is non-shelter services (in the lingo, “super-core services"). Price changes for these have been decelerating, and rose in the latest reading by less than anytime in the last two years.

Investors liked the CPI report because—even if prices for most items continue to rise faster than before the pandemic—the pace of these rises is slowing, albeit erratically.

Of course, economic data tells us about the past. Who can say these recent positive developments will continue? Fed Chairman Powell’s concern is that there are too many job openings for too few workers, and that the consequent wage gains will prevent inflation from decelerating further. Wages have been rising at a pace more consistent with 3-4% overall inflation, above the Fed’s goal of 2%.

But jobs openings are now falling. And recent research from the Federal Reserve Bank of Chicago shows that Powell may have it in reverse: wages don’t drive inflation, they respond to it. As Chicago Fed President Austan Goolsbee put it, when wages are stickier than overall prices (as they appear to be today), “watching wages to predict inflation is like looking for lightning when you hear the thunder.”

Reasonable people can disagree, but most of the data now suggest the remaining pockets of high inflation are thunder from a storm which has largely passed. If that’s true, by the end of the summer we are more likely to be talking about recession than inflation.