Abstract

Thomas D. Evans and Richard B. Tiller "Seasonal Adjustment of CPS Labor Force Series During the Great Recession"

The Current Population Survey (CPS) is a monthly household survey that collects information on labor force characteristics for the United States. The seasonally adjusted monthly levels of employment, unemployment and unemployment rate are key indicators of the health of the economy. Recent economic turbulence related to the Great Recession has raised public interest in how the CPS series are seasonally adjusted. The most widely used approaches to seasonal adjustment apply weighted moving averages to the original data to separate the seasonal from the nonseasonal components of the data. These weights may be derived from a model fit to each series or from a non-parametric method using pre-determined filters. These methods have been successful because of their flexibility in accommodating stochastic changes in seasonal patterns. The presence of stochastic seasonality, however, creates conceptual and statistical difficulties in separating seasonal from the nonseasonal components. The difficulties are compounded during periods of rapid economic change.