Comparing ARIMA Model-Based and Census X-11 Seasonal Adjustment

• Activity Number: 324
• Type: Topic Contributed
• Date/Time: Tuesday, August 5, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): William R. Bell*+
• Companies: U.S. Census Bureau
• Keywords: moving-average ; linear filter ; canonical seasonal adjustment ; seasonality ; forecasting

Abstract

Methods of seasonal adjustment based on moving-average filters were developed starting in the 1920s, and more or less culminated in the Census X-11 method in 1965. In 1976, Cleveland and Tiao showed that seasonal adjustment filters obtained from time series component models could approximate those of X-11. This soon led to development of the canonical ARIMA model-based approach to seasonal adjustment (Hillmer and Tiao 1982), and later to additional research comparing X-11 and model-based seasonal adjustment filters. We will review these developments, including some recent work that provides further insights into differences between ARIMA model-based and X-11 seasonal adjustment.