Monday, November 09, 2009

Seasonality....

Anonyman asks, do you understand this econ adjustment thingy? maybe you can explain it on yer blog...

He is referring to this article in the NYTimes...

Why, yes, yes I can. Suppose I go outside here in North Carolina on a day in January, and it's 65 degrees and sunny. I might say to the LMM, "Gosh, it's warm! This is the warmest day I can remember."

Now, suppose I go outside on a day in August. It's 68 degrees and sunny. I might tell the LMM, "Gosh, it's so cool today. This the coolest day I can remember!"

Why would I call a 65 degree day "warm" and a 68 degree day "cool"? It's obvious, almost silly: 65 degrees is a warm day in winter, and 68 degrees is cool day in summer.

But, in some absolute sense, 68 is warmer than 65. Yet was not the knight foresworn! (From "As You Like It," see below). Because the 65 is unusual warm direction compared to the mean temp in January, while 68 is unusual in the cool direction compared to the mean temp is August.

The point being that seasonality adjustments (to over-simplify) are calculated as deviations from the (monthly, seasonal, whatever) mean. In fact, the actual data are "adjusted" to ensure universal comparability, after taking seasonality out of the numbers. What THAT means is that you can get apparent gibberish like the lead 'graf in the Times article:

The economic reactions over the weekend to Friday’s employment report all started from the assumption that things grew much worse in October. The unemployment rate leaped to 10.2 percent from 9.8 percent. Another 190,000 jobs vanished. Actually, none of that happened.

In reality, the government report says unemployment rates remained steady at 9.5 percent. And the number of jobs actually rose, by 80,000. And the number of jobs for college-educated Americans rose more than in any month in the last six years....

So why is this the first time you’ve seen those better-looking numbers? It is because the government adjusted them before they were released. The adjustments are for seasonality. For some reason, October is the month with the largest seasonal adjustment down in jobs. So the increase in the unemployment rate does not reflect people actually losing jobs. It reflects the belief that seasonal factors should have added more jobs than they did.


One more shot at explanation: suppose the high temp in NC for each day in January stuck at 56 degrees. And suppose it stayed at 56 degrees for the daily high all through the summer. Would you say that the climate stayed the same? No, you would say that the climate cooled, dramatically, because you mentally "seasonally adjust" the numbers for ...well...seasons. Same with economics: what seems like good news in October is not good news because it takes great news in October to be good.

Finally, for the explanation of the Shakespeare reference:

AS YOU LIKE IT, Act I, Sc. II

TOUCHSTONE: Of a certain knight that swore by his honour they
were good pancakes and swore by his honour the
mustard was naught: now I'll stand to it, the
pancakes were naught and the mustard was good, and
yet was not the knight forsworn.

CELIA: How prove you that, in the great heap of your
knowledge?

ROSALIND: Ay, marry, now unmuzzle your wisdom.

TOUCHSTONE: Stand you both forth now: stroke your chins, and
swear by your beards that I am a knave.

CELIA: By our beards, if we had them, thou art.

TOUCHSTONE: By my knavery, if I had it, then I were; but if you
swear by that that is not, you are not forsworn: no
more was this knight swearing by his honour, for he
never had any; or if he had, he had sworn it away
before ever he saw those pancakes or that mustard.


So, maybe the answer is that the economists swear by their honor, though they never had any, and that's why seasonality makes sense!

1 comment:

Tom said...

Thank you for the lesson, Dear Mungowitz. Your expository skills delight me.