For a while I've been thinking that come January, we will be over the hump of the sharp CPI declines in 4q 2008. At that point the financial press won't be able to point to year/year declines in prices, and I will be very curious to see if they finally admit we are not stuck in a deflationary trap.

But it occurred to me tonight that the December announcement might actually go positive, since a bunch of the decline happened in October and November of 2008.

I checked the numbers, and here's what I found: The arithmetic mean (i.e. what the average person means by "average") of one-month non-seasonally adjusted CPI increases in January 2009 through October 2009 was 0.28%. So let's assume that the November CPI (which comes out in December) will be 0.28% higher than the October CPI, and that the December CPI (which comes out in January) will be 0.28% higher again.

If that happens, then on December 16, the BLS will announce that the November 2009 CPI is 1.02% 2.1% higher than the November 2008 CPI. In other words, the media will have to report that the monthly figure rose 28 basis points, and year/year prices are up 2.1%. Now maybe they'll use some seasonal jujitsu, or switch to a goods basket that excludes turkey and Christmas lights, but I think it's going to be hard for them to say, "At some point down the road, Bernanke will have to think about raising interest rates to contain inflation."

Then, come mid-January, the press will report on the December 2009 CPI, which again (in our scenario) will be 28 basis points higher than the last month's. This time, however, year/year CPI will have jumped by 3.4%. Then the analysts will probably freak out and say, "Oh my gosh!! Inflation has risen by more than 50% in one month! Aaaaagh!! It must have been consumers going crazy for Kwanzaa!!"

Then it's just a matter of time before we're all using ameros.