Large NSO retail sales figures revision

There's been a sizable revision of last years non food retail sales figures changing the annual rate of growth by up to +600%. I'm aware that the past couple of months are often revised, but I've never noticed an entire year being revised. Perhaps the modified the methodology for 2006 and have changed the 2005 figures to allow for comparisons to be made. Perhaps their forecasts are routinely 600% out and I've never noticed.

Any thoughts ?

Original Figures issued December 2005

Annual Index change Jan-04 127.6 Feb-04 127.8 Mar-04 129.4 Apr-04 130 May-04 131.1 Jun-04 131.7 Jul-04 131.4 Aug-04 131.9 Revised Figures issued Jan 2006 Sep-04 133.2 Change in Oct-04 132 Annual Index the rate Nov-04 132.1 Index change Revisionof growth Dec-04 129.9 Jan-05 129.7 1.6% 130.3 2.1% +0.6 28.6% Feb-05 130.3 2.0% 130.9 2.4% +0.6 24.0% Mar-05 130.8 1.1% 131.4 1.5% +0.6 42.9% Apr-05 130.7 0.5% 131.3 1.0% +0.6 85.7% May-05 130.6 -0.4% 131.2 0.1% +0.6 -120.0% Jun-05 132.5 0.6% 133.1 1.1% +0.6 75.0% Jul-05 131.4 0.0% 131.8 0.3% +0.4 Aug-05 132.7 0.6% 133.3 1.1% +0.6 75.0% Sep-05 133.3 0.1% 133.9 0.5% +0.6 600.0% Oct-05 133.5 1.1% 134.2 1.7% +0.7 46.7% Nov-05 134.6 1.9% 135.7 2.7% +1.1 44.0%

"The revisions result mainly from the seasonal adjustment process. The latest data, i.e. for Dec 2005, mean that we can update our estimates of the seasonal impact that December has had in recent years. The relatively strong Dec in 2005 means that we see the relatively weak Dec in 2004 in a slightly worse light. Consequently the seasonally adjusted estimates for EAPV for Dec 04 are adjusted downwards. The current revisions policy means we don't show the revised Dec 04 figure but we do reflect the new change from Dec 04 to Jan 05, i.e. the monthly change between Dec 04 and Jan 05 has been revised up by 0.6 % points."

Look like I should be using non seasonally adjusted prices for longer term comparisons.

Daytona