We recognize revenue when services are rendered. Because we collect payment in advance of services being performed, we defer the revenue until the services are performed. This is classified as a liability, Deferred revenue-prepaid postage, on our balance sheets. In Quarter III of 2008, we changed our methodology used to estimate the deferred revenue for prepaid postage for stamps. We further refined this methodology in Quarter IV of 2009 to reflect additional information in customer usage patterns of both Forever and denominated stamps demonstrated by newly available data.
Both updates were made necessary because the introduction of the Forever Stamp in April 2007, combined with the May 2008 price increase, resulted in a change in consumer behavior regarding the purchase and usage of stamps that was not directly measurable using the prior estimation techniques. In 2008, we enhanced our approach that more accurately captured trends in stamp usage. In 2009, based on newly available data, we refined this estimation model to allow us to better estimate consumer behavior patterns for the purchase and use of stamps. The 2008 change to an enhanced estimation technique and the 2009 refinement are both considered to be changes in accounting estimate under GAAP.
As required by ASC 250 (formerly FAS 154, Accounting Changes and Error Corrections), the impacts of the changes were recorded in the Quarter which the estimations were made: Quarter IV of 2009 and Quarter III of 2008.
For the year ended September 30, 2009, we increased the stamp portion of the deferred revenue-prepaid postage liability by $846 million, $655 million of which is considered a change in estimate that is attributable to changes in consumer behavior that were not identifiable based on data available previously.
For the year ended September 30, 2008, we increased the stamp portion of the deferred revenue-prepaid postage liability by $477 million, $230 million of which is considered a cumulative change in estimate and $247 million of which is attributable to changes in consumer behavior during the last two quarters of the year.
