Residual Seasonality in U.S. GDP Data

21 Pages Posted: 10 Nov 2016

See all articles by Keith R. Phillips

Keith R. Phillips

Federal Reserve Banks - Federal Reserve Bank of Dallas

Jack Wang

Federal Reserve Banks - Federal Reserve Bank of Dallas

Date Written: 2016-11-01

Abstract

Rudebush et al (2015a, b) and the Bureau of Economic Analysis find the presence of residual seasonality in the official estimates of U.S. real gross domestic product (GDP). Directly seasonally adjusting official seasonally adjusted GDP, which we refer to as double seasonal adjustment, could revise the first quarter growth in the past several years upward by an average of about 1.5 percentage points. The presence of residual seasonality can significantly distort current analysis of national and regional economies. In this paper we look more closely at the U.S. GDP data and study the quality of the seasonal adjustment when it is applied to data that has already been indirectly seasonally adjusted. We find that double seasonal adjustment can lead to estimates that are of moderate quality. While the optimal method would be to directly seasonally adjust the aggregate not seasonally adjusted data, if this is not possible, double seasonally adjusted data would likely lead to better estimates.

Suggested Citation

Phillips, Keith R. and Wang, Jack, Residual Seasonality in U.S. GDP Data (2016-11-01). Available at SSRN: https://ssrn.com/abstract=2866843 or http://dx.doi.org/10.24149/wp1608

Keith R. Phillips (Contact Author)

Federal Reserve Banks - Federal Reserve Bank of Dallas

Jack Wang

Federal Reserve Banks - Federal Reserve Bank of Dallas

2200 North Pearl Street
PO Box 655906
Dallas, TX 75265-5906
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
31
Abstract Views
255
PlumX Metrics