On Sunday, Zero Hedge ran a piece by Nicholas Colas of BNY in which Colas stated that federal withholding-tax receipts were down 4% year-over-year for the month of February (through the 22nd). And he drew the exact wrong conclusion. February is in the books now and receipts were actually up 3.8%. See for yourself. First, look at the final total from February 2010 (click image to enlarge):
Inside the red box, you see $139,589, right? Now, let’s look at the final report for February 2011 that was released today:
Inside the red box, you see $144,891. And that is $5,302 higher than $139,589. And since the numbers are expressed in millions, that’s a $5.3 billion beat.
That is flat-out year-over-year growth of 3.8%.
Colas didn’t even get the sign of the number right! Ridiculous!
And there was a huge tax-cut too, right? The only rational, objective way to interpret this data is that the economy has been creating jobs at a very healthy clip. If you like, you can change “creating” into “re-instating jobs lost during the Great Recession”, but the idea that the economy is contracting, stalling, or even slowing down just doesn’t fly. As a matter of fact, things are likely accelerating, though we can’t tell for sure because the tax-cut has clouded this data.
So why did Colas get it so horribly wrong? He made a couple of rookie mistakes. First of all, he used a period that was too brief. Paychecks go out every other Friday, monthly, or twice-monthly. When comparing two periods of withholding data, you absolutely must use periods that contain equivalent pay-cycles.
Colas’s second rookie mistake was to synchronize to the calendar rather than the work week. February 2010 began on a Monday, while February 2011 began on a Tuesday. That will often make a big difference.
So, Colas didn’t use enough data, and the periods he compared were not equivalent. He calculated the right number, but for non-comparable data. Here is what Colas did:
Here is the correct way to handle this sensitive data:
Notice how I grouped the data into work-weeks (#5 through #8 of the year). Also notice that I made sure to have a complete pay-cycle.
That’s how it’s done.
Going by work-weeks, we come up with a gain of 1.71%. And I believe that is the correct interpretation. Calendar February was up 3.8%, but that should be taken with a grain of salt because the calendar was a bit out of sync.
In any case, the fact that the data is higher after a large tax-cut is pretty miraculous.
As to the markets, they may well continue to react to events in the Middle East rather than to our expanding economy.