The Trump administration has proposed changing how inflation is measured when setting federal poverty thresholds because it believes that the current measure, the Consumer Price Index for All Urban Consumers (CPI-U), overstates inflation.

The administration is seeking and considering public input on an alternative measure, the Chained Consumer Price Index for All Urban Consumers, commonly known as the “chained CPI.” Some economists argue that the chained CPI provides a more accurate measure of inflation because unlike the CPI-U, it accounts for shifts in consumer behavior as prices for goods and services change.    

As I've previously written on Urban Wire, switching the inflation measure from the CPI-U to the chained CPI would result in slower inflation rates from year to year. The differences between the two inflation measures would be minimal at first but would compound over time.

What would this mean for low-income people?

Fewer low-income Americans would be included among those living under the federal poverty level and fewer would qualify for programs that use poverty guidelines (which are based on federal poverty thresholds) to determine eligibility.

One program that relies on poverty guidelines to determine eligibility is the Supplemental Nutrition Assistance Program (SNAP), the nation’s primary food assistance program, which serves 40 million people per month. In our new brief, we use the Urban Institute’s Analysis of Transfer, Taxes, and Income Security microsimulation model and 2016 American Community Survey data to estimate the number of people who would ultimately lose SNAP benefits if the poverty guidelines were based on federal poverty thresholds adjusted for inflation using the chained CPI.

We find that the following would have been true in 2016:

• About 579,000 SNAP recipients would have been ineligible for SNAP if the chained CPI had been the inflation measure used to adjust federal poverty thresholds for the previous 15 years. Among those recipients, 242,000—or about 42 percent—would have been children.
• The number of recipients losing SNAP eligibility would increase over time. Had the chained CPI been used for 5 years before 2016, 104,000 SNAP recipients would have been ineligible, and if it had been used for 10 years, 245,0000 recipients would have been ineligible. 
• Had the chained CPI been used for the previous 15 years, just over 240,000 SNAP households would have been ineligible in the average month in 2016, including nearly 50,000 households with a person age 60 or older, more than 20,000 households with a person with a disability, and more than 118,000 households with at least one child.
• The number of households that would have been ineligible in 2016 also varies by state, with more populous states experiencing the largest reductions in eligibility. Had the chained CPI been used for the previous 15 years, 24,000 households in New York and 15,000 households in California would have been ineligible.

SNAP isn’t the only safety net program that would be affected if we changed how we measure inflation. Recent research from the Center on Budget and Policy Priorities shows that in 10 years after adopting the chained CPI, more than 300,000 children would lose eligibility for health insurance coverage under Medicaid and the Children’s Health Insurance Program.

Shifting our inflation measure from the CPI-U to the chained CPI might seem like a technical, innocuous decision, but its effects would compound over time. Adjusting federal poverty thresholds using the chained CPI would erode eligibility for SNAP and other vital assistance programs for many vulnerable families.

The gains we have made in reducing food insecurity and uninsurance rates through the safety net could also be lost. Ultimately, the chained CPI doesn’t produce a more accurate measure of poverty or help us better understand the resources a family needs to exit poverty today.