For investors closely monitoring the stock market, the first Friday of every month is labor day – small “l”. On this day the Bureau of Labor Statistics (BLS) releases its monthly non-farm payroll report.

And last Friday was not a happy labor day.  Last Friday the number came in at 38,000, quite a bit shy of the 164,000 predicted by economists. The S&P 500 dipped almost 1% and the dollar had one of its worst days in almost seven years. Investors’ short-term, gut reaction: a dramatic economic slowdown is in the offing.

Let’s all take a breath. This report, as most estimated data sets, is based on squishy, not hard, data. So we may see huge gaps between estimated and actual trends. In fact, the labor report is almost always revised at a later date, sometimes by a lot.

After digging into the details we found a few devils, and we’re convinced that investors should take this latest statistic with a 10-pound grain of salt.

Unfortunately, Fed Chair Janet Yellen may not read this column (or at least not take it to heart). The labor day report is important because the Fed routinely says that non-farm payroll  is the one statistic it weighs most heavily when determining the path of interest rates. Yellen specifically calls out the non-farm labor report as a statistic that shapes its perception of economic growth.

Devilish Details

We dug into the BLS website to understand how this data is collected. Surprise: it doesn’t come from the Department of Unemployment, with its access to hard data on the number of unemployed citizens.

Instead, it’s done by surveys. Two surveys are sent into public domain to harvest information. One survey is sent to companies (the Establishment survey) and another to households (the Household survey).

The Establishment survey is done by computer-assisted telephone interviews (i.e.; press 1 if you’ve hired more than one person last month) and by web questionnaires. About 400,000 firms are queried each month.

This Establishment survey seems most reliable, but it ignores a growing and increasingly influential pool of self-employed and contract employees. As this population expands, the correlation between the Establishment survey and the health of the economy is diverging.

The Household survey is a wild card. This is a phone survey of 60,000 households. The questions stay the same each month, but as with any human process, it is open to error. Weather, health and personal situations can skew the data wildly. Many of the questions allow for ample interpretation. For one household member glancing through the help wanted ads might qualify as “looking for work,” but another might consider sending out resumes as actively seeking employment.

Buy the Dips

A chart of the monthly non-farm payroll data resembles a treacherous mountain path of steep spikes and jaw dropping gaps. We advise investors to watch the blended trend line, which right now illustrates an expanding workforce and healthy economic growth.

So instead of being spooked along with the herd after a scary monthly report, keep your eyes focused on the long-term trend and have your buy list ready.