Sections

Commentary

The Hutchins Center Explains: The Chained CPI

Michael Ng and
Michael Ng Former Research Analyst - Hutchins Center on Fiscal and Monetary Policy, The Brookings Institution
David Wessel

December 7, 2017

Editor's note:

This post was most recently updated on June 9, 2023.

provision in the tax bill signed into law in 2017 focuses attention on the measure the government uses to adjust tax brackets and other elements of the tax code for inflation. The bill substitutes a measure called “the chained consumer price index” (C-CPI-U) for the one previously mandated by law, the consumer price index (CPI-U), which we’ll call the primary CPI to avoid confusion. Here’s what this is all about.

What’s the difference between the primary CPI (CPI-U) and the chained CPI (C-CPI-U)?

Both indices are designed by the Bureau of Labor Statistics (BLS) to measure price changes faced by urban consumers. Each of the indices measures the prices of about 80,000 items, for both goods and services, each month across the country.

The BLS uses a fixed basket of goods and services based on a survey of 7,000 American families known as the Consumer Expenditures Survey . The survey determines what goods and services go in the basket and how much weight each should get in calculating the overall change in prices. For example, most people spend more money on cell phones than they do on landline phones, so changes in prices in the former are weighted more heavily. The market basket on which the primary CPI is based is updated every two years.

According to many analysts, the primary CPI overstates increases in the cost of living because it doesn’t fully account for the fact that consumers change their buying patterns when the price of one item goes up or the price of another goes down. When the price of beef increases, for example, consumers tend to buy less beef but more chicken; their total grocery bill doesn’t increase as much as it would have if they hadn’t changed what they purchased. When this ability to substitute for lower-priced products is ignored, the burden of inflation on consumers is overstated. (In 1999, the BLS did make a change to account for substitution within categories, such as consumers buying more medium-sized eggs and fewer large eggs when the latter go on sale, but not for between categories, such as beef and chicken.)

The chained CPI, however, better reflects changes in consumer buying patterns because it uses information about what people buy in the period before and the period after a price change. Essentially, the BLS calculates one measure of inflation that uses the first period basket and another measure of inflation that uses the second period basket (which might include different items because of price changes—more chicken and less beef, for example)—and then takes the average. It does this every month, and creates an index that “chains” these changes from month to month. This more frequent consumption data allows the index to reflect changes in consumer buying patterns between months and between categories.

How much difference does this make?

In the long-term, it can make a big difference. In general, because the chained CPI controls for the fact that people substitute away from items with large price increases, it tends to be a bit lower than the primary CPI. Since 2000, the primary CPI has risen by 45.7 percent and the chained CPI by only 39.7 percent, a difference of 6 percentage points.  This illustrates how seemingly small differences add up over time.

CPI - revised

So what does this have to do with taxes?

By law, tax brackets are adjusted each year to account for changes in the cost of living. For example, the Internal Revenue Service adjusted the income threshold for the top 39.6 percent marginal tax rate from $415,050 in 2016 to $418,400 in 2017. These adjustments prevent taxpayers from being pushed into higher tax brackets when their incomes rise just enough to keep up with inflation.

By law, the brackets and other elements of the tax code were previously adjusted using the primary CPI (CPI-U). The tax bill passed in 2017 by Congress changed the law to use the chained CPI (C-CPI-U). Because the chained CPI climbs more slowly than the primary CPI, the tax bracket thresholds will increase by smaller amounts each year.  More people will be pushed into higher tax brackets than they would have under the old law. In short, Americans will pay more taxes over the next ten years, and tax revenues will be about $134 billion higher than they would have been without this change in the law, according to the congressional Joint Committee on Taxation.”

What about the inflation adjustment for spending programs?

Some other government spending programs, notably Social Security, are also indexed for inflation but don’t use the chained CPI. The pending tax bill wouldn’t change that. However, in his 2014 and 2015 budgets, President Obama proposed using the chained CPI for both taxes and spending to help reduce the budget deficit. Congress didn’t adopt his proposal.

What other parts of the government are using chained CPI?

In June 2023, the Census Bureau said it is adopting the chained CPI to adjust historical estimates of median income for inflation going back to 2000. Previously, the Census used a variant of the primary CPI, retroactively adjusting the inflation measure only for changes in methodology. Because the chained CPI estimates lower rates of inflation, using that measure will show inflation-adjusted median income growing faster over time than in previous official reports, but only slightly faster. For instance, Census initially reported that inflation-adjusted median income grew 6.8% in 2019 and fell 2.2% in 2020, the first year of the pandemic. Had it been using the chained CPI, it would have reported that median income grew 7.2% in 2019 and fell 2.1% in 2020.

What are the drawbacks to using the chained CPI?

The primary CPI is reported each month, and never revised. Not so for the chained CPI. The chained CPI rests on data released each month on consumer buying patterns, and these data are revised several times a year. Therefore, the chained CPI is also revised several times; a final reading is posted between 10 and 16 months after the initial release. The CBO has discussed this as a potential downside for using the chained CPI to index taxes and benefits.

Finally, it’s worth noting that the CPI, chained or not, is designed to capture the experience of the average household.  Its measure of changes in the cost of living likely differs from that experienced by particular individuals or families.

Authors