Monitoring CPS Seasonally Adjusted Series with an Eye to Recession Effects

Thomas D. Evans and Jennifer P. Oh

Abstract

Seasonal adjustment of national Current Population Survey (CPS) series is important to help understand the health of the U.S. economy. Concern is occasionally expressed in the news media that unusual events such as recessions can affect seasonal factors. The majority of outliers that occur in CPS series have known causes as to type and duration. For example, these effects are often due to known shifts in population controls, severe storms, or changes to the survey instrument. While it is unusual to try and handle other potentially distorting effects until there are many observations after the fact, it is useful to monitor for these potential outliers in real time. The monthly monitoring for CPS seasonally adjusted series is described below and includes numerous tables, diagnostics, and graphics, and real-time outlier analysis. Results are also presented for monitoring changes in the seasonal ARIMA coefficients over time. This battery of diagnostics should also give clues for any evidence of recession effects in the seasonal adjustment process.