A critique of the traditional composite index methodology

Article type: Research Article

Authors: Boldin, Michael D.^a

Affiliations: [a] Economy. Com, Inc., 600 Willowbrook Lane, Suite 600, West Chester, PA 19382-5500, USA. Tel.: +1 610 241 3335; Fax: +1 610 696 3836; E-mail: boldin@dismal.com

Abstract: The leading index receives attention because this summary statistic of a particular set of popular economic indicators is touted as a forward-looking and objective measure of economic activity. A review of its history and methodology, however, leaves much to be desired. It is well known that changes in the components have been responsible for substantial revisions in the leading index. Less known is the role of the methodology in these changes. This article critiques the methodology in this light and considers whether a consistent shying away from modern econometrics has produced excessive revisions in the leading index. The conclusion that the traditional composite index approach has serious, but correctable flaws should be of interest to those that want to improve the leading index or build similar economic statistics for subsectors of the US economy, US regions, or other countries.

DOI: 10.3233/JEM-1999-0170

Journal: Journal of Economic and Social Measurement, vol. 25, no. 3-4, pp. 119-140, 1999