U.S. flag

An official website of the United States government

Department of Labor Logo United States Department of Labor
Dot gov

The .gov means it's official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you're on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Bureau of Labor Statistics > Office of Survey Methods Research > Publications > Browse Research Papers

Assessing Nonresponse Bias in the Consumer Price Index Surveys

Asia Anuwa, Jenny Fitzgerald, Johanky Reyes, Alexander Traczyk, and Alice Yu

Abstract

The Consumer Price Index (CPI) estimates the change in prices over time of the goods and services U.S. consumers buy for day-to-day living. Prior to the COVID-19 pandemic, the outlet and quote repricing response rates for the CPI’s Commodities and Services (C&S) Survey were about 90 percent and 80 percent, respectively, and the unit repricing response rate for the CPI Housing Survey were about 70 percent. As brick-and-mortar stores shut their doors and the U.S. adopted social distancing to reduce the risk of spreading COVID-19 in 2020, the response rates for both surveys steadily declined. In December 2023, the outlet and quote repricing response rates for the C&S Survey were still only 80 percent and 70 percent, respectively, and the unit response rate for the Housing Survey was about 68 percent. To meet the Office of Management and Budget (OMB) standards and guidelines for statistical surveys regarding response rates, research was conducted using a three-phase approach to assess the effect of nonresponse bias on the C&S and Housing Indexes from January 2018 – December 2023. In the first phase of the study, sampling unit characteristics contributing to nonresponse at sample initiation were identified using Random Forest for the C&S Survey, and nonresponse adjustment cells were created using decision trees. In the second phase of the study, sampling unit characteristics contributing to nonresponse at repricing were investigated using Random Forests, Random Survival Forests, and Generalized Linear Mixed Model Trees for the Housing Survey. In the final phase of the study, counterfactual indexes were calculated to determine whether the level of nonresponse error in the price change estimates was significant by imputing price relatives for nonresponding units using models developed based on the results of the previous two phases.