In what sense is it a 1/4? Because when asked in censuses around twice as many people claim to be JWs than in the publisher count. Are you saying you know better than people themselves whether they are JWs when asked? JWs have one of the strictest methods of counting members of any religious group so even if they “overcount” by their own definition it is still a significant undercount by any comparison with the membership of other groups. And the difference is dramatic.
For example in 2010 JWs claimed 706,699 publishers in Brazil but in the census 1,393,208 identified themselves as JWs. That’s 1.97 times as many describes themselves as JWs than are counted as publishers.
By contrast in that same year Mormons claimed 1,138,740 members in Brazil but in the census only 226,509 described themselves as Mormons. That’s only 0.20 times the number of official Mormon members actually identified in the census.
These results are replicated over lots of countries over many years.
See more details in the article here: