Seasonal Adjustment and Modeling of Higher Frequency Time Series in UK Official Statistics During the COVID-19 Outbreak

• Activity Number: 199 - Novel Time Series Approaches for Official Statistics in the Time of COVID-19
• Type: Topic-Contributed
• Date/Time: Tuesday, August 10, 2021 : 1:30 PM to 3:20 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Duncan Elliott*
• Companies: Office for National Statistics
• Keywords: seasonal adjustment; high frequency; official statistics; time series

Abstract

From the beginning of the COVID-19 outbreak in the UK there has been an increased demand for timely and more frequent official statistics. The UK Office for National Statistics has responded to this increased demand with publications of higher frequency data sets, such as weekly estimates of selected labour market variables, and a faster indicators dataset, that includes a range of experimental data to assess the impact of COVID-19 on society and the economy. Seasonal adjustment is regularly used in official statistics to aid interpretation of movements in these time series using well established software such as X-13ARIMA-SEATS and JDemetra+. Neither software is currently designed to seasonally adjust time series at a higher frequency than monthly. This talk gives an overview of the variety of seasonal adjustment and time series modelling challenges that have been addressed during the COVID-19 outbreak, including examples of weekly and daily seasonal adjustment using experimental methods in rjdhf, an R package for testing methods for possible future releases of JDemetra+.