If you pay attention to the economic indicators that Wall St. waits for, you’ll have heard about the “jobs report” which reports on the number of jobs created over the last month in the U.S. Over the last 12 months, the jobs report has continued to look decent, and last month the jobs report showed strong growth, including stable employment in the home construction and financial sectors.

That last sentence should raise a red flag if you’ve been paying attention to the housing market, and Barry Ritholtz over at the Big Picture has been consistently complaining about a particular feature of the NFP (Non-farm Payroll, or jobs) reporting. I’ll leave you with two thoughts and then suggest you head over to his blog to get the big picture. I can’t explain it properly, as I’m less than an amateur here.

First thought, look at this picture:

The second thought is a quick explanation of that image. The bottom section is the number of jobs created by a statistical correction the BLS uses to estimate jobs created or lost by firms too small to make the survey. In other words, it’s an modeled estimate, not a number based on surveying actual firms out in the real world. The red section is the portion of the jobs number that month that comes from that, real world survey. The other important point here is that this report is supposed to be primarily based on that real world sampling.

Think about that, and then go over to Barry Ritholtz’s blog to get a better explanation and more data. This stuff is important because it drives financial modeling (among many other things). If this data is flawed, and assuming that the gap between the model and reality is getting wider (as the chart indicates), then we’re going to be in for a significant correction later as policies will need to adjust sharply. That long decline will feel like a sharp cliff once policies have to adjust to real numbers.

At least, that’s my interpretation. Like I said, go read Ritholtz.