Measuring The Statistical Insignificance Of The Monthly Jobs Report (September 2007 Edition)

• Tweet

This morning, the Non-Farm Payrolls report showed a net loss of 4,000 jobs in the month of August, its first net decline since 2003.

Markets expected some weakness in the report, but not this kind of weakness.  Already, mortgage markets are up 31 basis points today. 

That's a swift, powerful reaction, especially considering that the unemployment rate remained unchanged at 4.6% and is viewed as "strong".

Today, mortgage rates are moving because the investors think the Fed now has an economic reason to lower the Fed Funds Rate later this month. 

But is the economics even there?  If we look deeper at the numbers, we can answer the question(s): What is the true significance of this morning's Non-Farm Payrolls report?  Is the mortgage market's reaction justified? 

Consider the following (subject to revision in October and November):

• 4,000 jobs were lost in August versus expectations of 110,000
• 24,000 fewer jobs were created in July than previously measured
• 57,000 fewer jobs were created in June than previously measured

Adding it up, today's actual news was that the number of working Americans was off by a measure of 195,000 against the total number of employed people of 146,000,000.  In percentage terms, the "surprise" represents 0.134% of the overall workforce.

Now, let's put that 0.134% adjustment in mathematical perspective:

• 0.134 percent is one foot in the height of the Golden Gate Bridge
• 0.134 percent is one person living in Foxfield, Colorado
• 0.134 percent is one double in the storied career of Charlie Hustle

Statistically, 0.134 percent is insignificant. 

And yet, mortgage rates are plowing lower today while economists chirp about dramatic economic weakness tied to the credit markets.  If I hear one more person say that Fed has to lower the Fed Funds Rate because of today's payroll report, I'll barf.

So, just like the last time I did a study like this, the market is reacting strongly to data because of its psychological implications, not because of a fundamental analysis.