Study of Volatility and Smoothing of Estimated Variances in the Employment Cost Index Program

• Activity Number: 465
• Type: Contributed
• Date/Time: Wednesday, August 9, 2006 : 2:00 PM to 3:50 PM
• Sponsor: Section on Survey Research Methods
• Author(s): Meghan S. O'Malley*+ and Daniell Toth and Chester Ponikowski
• Companies: Bureau of Labor Statistics and Bureau of Labor Statistics and Bureau of Labor Statistics
• Keywords: variance smoothing ; exponential smoother ; nonstationary time series ; seasonality ; Laspeyres index

Abstract

The Employment Cost Index (ECI), one of the key products produced by the National Compensation Survey (NCS) and conducted by the Bureau of Labor Statistics (BLS), is a Laspeyres index that provides quarterly and annual measures of the rate of change in compensation per hour worked including wages and salaries, and employer cost of employee benefits. Point estimates for ECI variances are produced using the balanced repeated replication (BRR) procedure and provided to data users as smoothed estimates in the form of five-year moving averages of these BRR variance estimates. This paper presents an analysis of the volatility of the estimated variances as a time series, discussion of the appropriateness of smoothing, development of exponential smoothing procedures taking into account specific properties of the ECI variance estimates, and a comparison of the results.