Seasonal Adjustment in the CPI

The Consumer Price Indexes program seasonally adjusts price index series that demonstrate significant seasonal patterns. Every year, series are analyzed with respect to whether there is evidence of a seasonal pattern that makes seasonal adjustment appropriate. In the past the series for owners' equivalent rent showed enough evidence of seasonal price patterns to qualify for seasonal adjustment. The seasonal pattern was very small, with seasonal factors close to 100.000. However, for the past three years the series has failed to show a significant pattern, and thus will not be seasonally adjusted in 2012. The series may be seasonally adjusted again in the future if it resumes showing a seasonal pattern. All seasonal adjustment status changes for 2012 are listed in Series changing seasonal status January 2007-December 2011.

Each year, with the release of the January CPI, seasonal adjustment factors are recalculated to reflect price movements from the just-completed calendar year. This routine recalculation may result in revisions to seasonally adjusted indexes for the previous five years.

Basic information on the use of seasonal adjustment may be found in our Fact Sheet on Seasonal Adjustment.

For a more technical discussion of seasonal adjustment methodology, see the following excerpts from the BLS Handbook of Methods:

Seasonal Data and techniques for Data for January 2007-December 2011, Issued February 2012

A Note on the Use of Seasonally Adjusted and Unadjusted Data

Because price data are used for different purposes by different groups, the Bureau of Labor Statistics publishes seasonally adjusted as well as unadjusted changes each month.

For analyzing general price trends in the economy, seasonally adjusted changes are usually preferred since they eliminate the effect of changes that normally occur at the same time and in about the same magnitude every year—such as price movements resulting from changing climatic conditions, production cycles, model changeovers, holidays, and sales.

The unadjusted data are of primary interest to consumers concerned about the prices they actually pay. Unadjusted data also are used extensively for escalation purposes. Many collective bargaining contract agreements and pension plans, for example, tie compensation changes to the Consumer Price Index before adjustment for seasonal variation.

Seasonal factors used in computing the seasonally adjusted indexes are derived by the X-12- ARIMA Seasonal Adjustment Method. Seasonally adjusted indexes and seasonal factors are computed annually. Each year, the last 5 years of seasonally adjusted data are revised. Data from January 2007 through December 2011 were replaced in January 2012. Exceptions to the usual revision schedule were: the updated seasonal data at the end of 1977 replaced data from 1967 through 1977; and, in January 2002, dependently seasonally adjusted series were revised for January 1987-December 2001 as a result of a change in the aggregation weights for dependently adjusted series. For further information, please see "Aggregation of Dependently Adjusted Seasonally Adjusted Series" (PDF).

Effective with the publication of data from January 2006 through December 2010 in January 2011, the Video and audio series and the Information technology, hardware and services series were changed from independently adjusted to dependently adjusted. This resulted in an increase in the number of seasonal components used in deriving seasonal movement of the All items and 54 other lower level aggregations, from 73 for the publication of January 1998 through December 2005 data to 82 for the publication of seasonally adjusted data for January 2006 and later. Each year the seasonal status of every series is reevaluated based upon certain statistical criteria. If any of the 82 components change their seasonal adjustment status from seasonally adjusted to not seasonally adjusted, not seasonally adjusted data will be used in the aggregation of the dependent series for the last 5 years, but the seasonally adjusted indexes before that period will not be changed. Note: 37 of the 82 components are not seasonally adjusted for 2012.

Seasonally adjusted data, including the All items index levels, are subject to revision for up to five years after their original release. For this reason, BLS advises against the use of these data in escalation agreements.

Effective with the calculation of the seasonal factors for 1990, the Bureau of Labor Statistics has used an enhanced seasonal adjustment procedure called Intervention Analysis Seasonal Adjustment for some CPI series. Intervention Analysis Seasonal Adjustment allows for better estimates of seasonally adjusted data. Extreme values and/or sharp movements that might distort the seasonal pattern are estimated and removed from the data prior to calculation of seasonal factors. Beginning with the calculation of seasonal factors for 1996, X-12-ARIMA software was used for Intervention Analysis Seasonal Adjustment. For more information, see "Improvements to CPI Procedures for Intervention Analysis Seasonal Adjustment" (PDF), which was originally published in the the December 1996 CPI Detailed Report.

For the seasonal factors introduced in January 2011, BLS adjusted 29 series using Intervention Analysis Seasonal Adjustment, including selected food and beverage items, motor fuels, electricity and vehicles. For example, this procedure was used for the Motor fuel series to offset the effects of events such as damage to oil refineries from Hurricane Katrina.

For a complete list of series that used Intervention Analysis Seasonal Adjustment, see "Intervention Analysis Seasonal Adjustment" (PDF).

For additional information on seasonal adjustment in the CPI, please write to the Bureau of Labor Statistics, Division of Consumer Prices and Price Indexes, Washington, DC 20212 or contact David Levin on (202) 691-5261 or by e-mail at levin.david@bls.gov. If you have general questions about the CPI, please call our information staff at (202) 691-7000.

 

Last Modified Date: February 15, 2012