It is from this document that certain critics erroneously concluded that 95% of new AA members leave at the end of the first year. What the statistics in this document actually say is that 5% of the members in their first year of attendance of the sampled meetings are in their 12th month of attendance.. If this document had been better written and more readable then the erroneous conclusion would not have been drawn. We cannot tell from this document how many AA newcomers leave at the end of the first year.
It is clear that the anonymous author of this document is quite unfamiliar with statistics and is incapable of manipulating them properly. At one point on page four this anonymous author erroneously concludes that over half of AA newcomers leave AA after three months. The author bases this on viewing data only for members in their first year of attendance--and it is of course impossible to draw a conclusion on how many members leave at the conclusion of three months unless one looks at the data for all members, not just those in the first year.
We must also point out the fact that the sampling of meetings discussed in this paper and in AA's Pamphlet P-48 "Alcoholics Anonymous 2007 Membership Survey" is highly biased because it includes only meetings which are officially registered at AA's General Service Office (GSO). The vast majority of AA newcomers are attending "in house" AA meetings at treatment centers or halfway houses. These "in house" meetings do not meet the criteria for being listed by the AA GSO because they charge a fee to insurance companies and they are not open to all seeking help.
AA General Service Office (2007). AA Pamphlet P-48 Alcoholics Anonymous 2007 Membership Survey.