The Bureau of Economic Analysis (BEA) is working on a multi-pronged action plan to improve its estimates of gross domestic product (GDP) by identifying and mitigating potential sources of “residual” seasonality. That’s when seasonal patterns remain in data even after they are adjusted for seasonal variations.
Each spring, BEA conducts an extensive review–receiving updated seasonally adjusted data from the agencies that supply us with data used in our calculation of GDP. Most of the data the feeds into GDP is seasonally adjusted by the source agency, not BEA. At the same time, BEA examines its own seasonal factors for those series that BEA seasonally adjusts itself. All that work takes place in preparation for BEA’s annual revision to GDP and its major components, which will be released on July 30.
As a result of this ongoing work, BEA is aware of the potential for residual seasonality in GDP and its components, and the agency is looking for ways to minimize this phenomenon.
• One of the areas we’re currently reviewing is possible residual seasonality in measures of federal government defense services spending. Initial research suggests that the first and fourth quarter growth rates are lower on average than those of the third and second quarters. BEA is developing methods for addressing what it has found.
• Time frame to implement: Improvement will take place with the release of second quarter GDP on July 30. Period covered: 2012, 2013, 2014, and forward.
• BEA also will begin adjusting certain inventory investment series that currently aren’t seasonally adjusted.
• Time frame to implement: Improvement will take place with the release of second-quarter GDP on July 30. Period covered: 2012, 2013, 2014, and forward.
• Also as part of this year’s seasonal adjustment review, BEA is planning to seasonally adjust a number of series from the Census Bureau’s quarterly services survey that now have sufficient time spans to which seasonal adjustment techniques can be applied. Currently, these series are smoothed using a four-quarter moving average to attempt to smooth out seasonal trends in the data. While BEA’s review had not identified residual seasonality in the PCE services estimates, applying statistical seasonal adjustment techniques to these indicators will improve the accuracy of the underlying trends in PCE estimates.
• Time frame to implement: Improvement will take place with the release of second quarter GDP on July 30. Period covered 2012, 2013, 2014, and forward.
• BEA will review all series entering the GDP calculations to identify, and where feasible, mitigate any residual seasonality within its existing seasonal adjustment methodologies.
• Time frame to implement: Review will take place with the release of second-quarter GDP on July 30. Period covered: 2012, 2013, 2014, and forward.
• Longer term–beyond July 30–BEA will continue looking at components of GDP to determine if there are opportunities to improve seasonal adjustment methodologies. Should BEA identify other areas of potential residual seasonality, BEA will develop methods to address these findings. If research suggests that residual seasonality originates with already seasonally adjusted source data, BEA will work alongside its source data agencies to determine the appropriate course of action.
Additional information will be available in an upcoming article in BEA’s Survey of Current Business that’s slated to be published in mid-June.