Edward Lotterman
Measuring key aspects of a nation’s economy is important, but the metrics used are often misunderstood by many. Quantifying output of goods and services, price levels, and labor use provides actionable information that can help individuals, business managers and government officials make better decisions
Unfortunately, history shows indicators are subject not only to benign misinterpretation, but to use in overt demagoguery. Measures of inflation and deflation, changes up or down in the general price level, are the most misrepresented and hence mistrusted — especially when the data makes political agendas inconvenient. Nothing will ever change that.
Yet this data’s veracity is essential to good policy making. So the better that people of good will and open minds understand, and trust, what is going on, the better our collective democratic self-governance.
Price indexes are the best known indicators of our economy and the oldest. It was out of concern for managing the productivity of a rapidly growing nation that Congress established a Bureau of Labor Statistics in 1884. Intended to improve understanding of the numbers and situations of workers, it reached its current status as an agency within the new Department of Labor in 1913.
Four years later, with World War I raging, prices of food and other goods rose rapidly. As was the case before the 2024 election, concern in 1917 that such inflation, especially for food, was outstripping workers’ pay, and this became a red-hot political issue. This led the Bureau back then to begin measuring overall price levels for consumers.
The first scientifically designed price survey was undertaken in 1919 and done at monthly intervals thereafter. Major overhauls took place over time, each incorporating new techniques developed by economic researchers and addressing unanswered issues that arose.
So the Consumer Price Index, the metric that most people think of when discussing “inflation,” whether in headlines or policymaking rooms, has been honed for 105 years. It is not perfect, but it takes into consideration myriad methodological issues never would occur to most of us. (Search the internet for “BLS CPI Press Release” for a wealth of information.)
Every month the Bureau uses statistical survey techniques to ascertain current prices of goods and services used by consumers, whether as individuals, families or in generic “households.”
This index computes the weighted average prices of about 90,000 items ranging from ground beef to shoelaces to piano lessons to canned tuna to oil changes. These items are the “market basket” that consumers fill. The prices of all these products are recorded, largely at the points of sale such as gas stations, supermarkets, retail stores or, increasingly, online.
“Weighted” means that in computing each month’s index, the importance of each price recorded depends on what fraction spending on the item is of total household spending. Bread, milk, chicken and beef, gasoline, work clothing, housing expenses and the like have large weights. Black pepper, movie tickets, snow shovels and hair brushes have small weights. The “weights” are the decimal numbers by which the price of each item is multiplied before being added into the overall index. The weights for all items in the “market basket” add up to one.
The sum of all these weighted prices was set as a base of 100 for the first survey. Then, at some future month, if the weighted average cost of the market basket was 18.3% higher than in the base period, the index for that month would be 118.3. If it rose to 121.7 a year later, the index number increase of 3.4 divided by the starting level of 118.3 gives a rise in overall consumer prices of 2.9%. That was inflation over the year.
The base period is periodically reset to 100 to facilitate quick mental comparisons — it is easier to see an approximate 3% increase when an index rises from 207 to 213 than from 759 to 782. But whenever the base period is re-zeroed, past numbers are re-calculated so that the relative magnitudes in the series remain the same.
Use the older series with a base of 100 in 1967 to compare bread prices in 1952 with those in 1978. Repeat using the newer series with its base of 100 equaling the average of prices over the years 1982 through 1984. You find the same percentage change in the bread prices. One can compare on a coherent scale from 1913 to the present.
The CPI number for December 2024 was 317.685. It had been 316.441 in November and 308.742 a year earlier in December 2023. The year-to-year increase of 8.943 in the index represents a 2.896% increase in prices over 12 months. Rounded to 2.9%, this was the inflation number most commonly reported in the media.
The 12-month comparison can be misleading, however, because increases over the most recent month or few months can be higher or lower than over the full year average. Wise observers who want to know what is going on right now also calculate what would happen over 12 months if the increase over the most recent one or two or three months kept on for a year.
In the latest report, the November-to-December increase is 0.39%. If that monthly rate continued for 12 months, the cumulative increase would be 4.8%. If one takes the increase from September to December and annualizes it, the rate of change is 3.9% a year. So the rate of inflation in the most recent month or months is higher than it had been over 12 months.
So to fully understand how prices are changing, one needs to look at both the change from a year ago to now and what the current rate is converted to an annual one. The media seldom do this clearly. That causes confusion in the general public, even if not among financial market analysts.
There is another source of confusion, one often viewed as an attempt to mislead the public. Food and energy prices are more “volatile” than those for clothing, hardware, professional services and so on. They go up and down more.
Issues like bird flu or blockage of the Suez Canal can make prices of eggs and chicken or gasoline spike in any one month. Anyone who needs to make big decisions based on inflation rates — such as a Federal Reserve board member — wants to understand the underlying trend for less-affected items as well as the overall rate. Thus a “core” rate for “all items less food and energy” is calculated.
The two series, “headline” and “core” complement each other. Neither is definitive. Over time the two show nearly identical average changes. But the announcement of the core rate leads millions to mistakenly believe that “the government doesn’t include food and gasoline in the inflation numbers.” Wrong! It does indeed include them. The “core” number simply adds information to the full or “headline” number.
Another agency, the Bureau of Economic Analysis in the Department of Commerce, tabulates Gross Domestic Product. This is the value of all goods and services produced. To compute values, they must know prices of output. That data is collected directly from sellers of goods and services rather than by surveying items on store shelves as for the CPI. So in recent decades, the “Personal Consumption Expenditures” price index is tabulated and published as part of the overall GDP estimation process. A “core” PCE, without food and energy, is part of the release.
The PCE index differs from the CPI in ways that are small but important for some analyses. In 2000, the Fed announced it was using the PCE index rather than the CPI as its primary indicator. That change also gave rise to much conspiracy theorizing about government skullduggery to deceive the public. But like the full CPI versus the “core CPI,” PCE index changes are lower than those for the CPI in as many months as they are higher. Over time, the two end up being nearly equal.
Doubters will always doubt. Efforts to quantify complex variables like output or price levels in a nation’s economy inherently will be less precise than measuring air temperature or rainfall at one location. Anyone using price index information to make decisions must keep that in mind. But the indices are produced honestly with as sophisticated a set of survey and tabulation techniques as possible given the state of the art.
The real problem with the CPI and the PCE is the degree to which financial markets hang on how new price level numbers will influence Fed policy decisions. But that is a complex topic in itself.
Related Articles
Real World Economics: Steelmaking faces clash of new tech, obsolete values
Real World Economics: Steeling ourselves for a changing industry
Real World Economics: Enter the new year with a dose of knowledgeable skepticism
Real World Economics: Eliminating Social Security’s ‘windfall elimination’ would be an injustice
Real World Economics: Powers-that-be should heed the warning signs
St. Paul economist and writer Edward Lotterman can be reached at stpaul@edlotterman.com.
Leave a Reply