The Bureau of Labor Statistics’ monthly survey of employers is a popular report that investors use to gauge the strength of the labor market. However, the numbers can be way off from reality.

For instance, the bureau releases a quarterly report, titled “Business Employment Dynamics” that comes from state unemployment records. The latest report was released for the 3rd quarter of last year and showed that only a net 19,000 private sector jobs were added to the economy.

On the other hand, the widely reported monthly survey concluded that private-sector employment grew by 498,000 jobs — a healthy number considering the economy grew by only 2% during that period.

A big difference was in construction employment where the monthly survey determined that 34,000 net-jobs were added compared to a loss of 77,000 jobs in the quarterly study.

What’s causing the significant discrepancy? One reason may be that the bureau uses a “birth/death” model to estimate the change in employment from the launching and demise of businesses. Eventually, the monthly numbers will be revised to reflect the results of the quarterly survey, but that won’t be done until data is available for the 4th of last year and the 1st quarter of this one.

I tend to ignore the monthly jobs data because it is often subject to large revisions much later on, at which point the numbers are useless as a forecasting tool since employment is a coincident indicator of economic growth.