Indexes of Economic Indicators: What Can They Tell Us about the New England Economy
Clayton-Matthews, Alan, Kodrzycki, Yolanda K., Swaine, Daniel, New England Economic Review
London: The University of Chicago Press.
Watson, Mark W. and Robert F. Engle. 1983. "Alternative Algorithms for the Estimation of Dynamic Factor, MIMIC, and Varying Coefficient Regression Models." Journal of Econometrics, vol. 23, August, pp. 385-400.
RELATED ARTICLE: Solving Unobservable Variable Time-Series Models
Dynamic factor analysis and the Kalman filter involve different algorithmic methods to estimate models that are intimately related.(14) As Engle and Watson (1981) and Watson and Engle (1983) point out, a wide spectrum of factor analysis models are special cases of the so-called state-space model, Each month, the Boston Federal Reserve publication New England Economic Indicators reaches about 6,000 subscribers. Many of these readers undoubtedly want to know: "How is the New England economy doing?" In response, Indicators offers a comprehensive compilation of timely data. Yet no one of these indicators can possibly provide an overall assessment of the New England economy. Nor can any other individual piece of information offered by other regional or state agencies.
This article explores the development of composite coincident indexes summarizing the condition of the economy of New England and its six states. While composite indexes have long been used to analyze the national economy, they may be needed even more for regions because of the lack of current comprehensive measures of state activity, as well as the frequent lack of clarity in indicators that are available. The article discusses two approaches to constructing composite indexes, the traditional averaging method used by the U.S. Department of Commerce and a latent variable method advanced by two academic researchers, James Stock and Mark Watson, and recently applied to states in a study conducted at the Federal Reserve Bank of Philadelphia. A general model for deriving a coincident economic index is described, along with its relationship to the Stock-Watson and Philadelphia Fed research. The Philadelphia Fed specification is then applied to the New England states in order to measure how regional business cycles have compared with national cycles. In addition, an alternative composite coincident index is constructed for Massachusetts, using as inputs newly developed data based on state tax collections, which have the advantage of being quite timely and not subject to revisions (in contrast to more commonly used indicators).
Composite indexes of U.S. economic indicators have a long history. Originally developed a half century ago by researchers at the National Bureau of Economic Research, a private organization, such indexes are now issued monthly by the U.S. Department of Commerce. The government's index of leading indicators gets widespread attention when it is released. Less noticed, but released just as frequently, are the index of coincident indicators, which measures the current state of the national economy, and the index of lagging indicators.
Composite indexes are attractive because of their simplicity and compactness. Interest in economic trends is widespread, but most people have neither the time nor the inclination to try to interpret each individual piece of economic data that becomes available. Furthermore, even for professional observers of the macroeconomy, individual indicators may be problematic. Any particular economic series may give an ambiguous signal, or it may contradict another piece of data.
National composite indexes (and individual indicators) may be misleading for regions because regional business cycles do not parallel national cycles exactly. For example, the strongest recoveries from the early 1980s recessions occurred on the East and West Coasts, and several interior states lost employment or showed virtually no gain in employment for several years after the national economy had picked up (Bradbury and Kodrzycki 1992). By contrast, in the upturn from the most recent national recession of 1990-91, it is widely acknowledged that the New England states and California were laggards. …