Improving CES Preliminary Employment Estimate through Statistical Model

The U.S. Bureau of Labor Statistics (BLS) releases estimates of total non-farm business payroll employment changes every month. These estimates have significant impact on US economic policy and financial market decision makers. However these estimates are produced based only on "preliminary" Current Employment Statistics (CES) survey results, final estimates are release two months later. In this paper, we develop a statistical model based on historical CES data with the following goals in mind: (1) validate factors that affect monthly employment changes, (2) measure the magnitude of these effects on monthly employment movement, (3) prove in part some underlying economic factors affecting national and regional employment changes. We then use prediction results from the model to produce improved preliminary estimates that is more accurate than the current preliminary estimates.