Page:American Journal of Sociology Volume 15.djvu/829

 THE RECENT CENSUS OF RELIGIOUS BODIES 815

J. E. Cutler, Western Reserve University I should like to ask Dr. Coe a question. It has been generally under- stood, I think, that the census of churches and church membership taken in 1890 was extremely defective and on the whole rather unreliable. Pre- sumably the statistics given in Census Bulletin 103 are more complete and much more reliable. I should like to ask Dr. Coe, if, in using these two sources as a basis for his paper, he gave any consideration to the fact that one was possibly more complete and more trustworthy than the other.

George A. Coe

Jn reply to a question whether the apparent increase in church mem- bers may not be due in part to improved methods in securing census returns. Professor Coe said in substance:

"It is difficult to say exactly what effect upon membership statistics these improved methods have had. Even if the tendency of such improvement in the methods of government were to enumerate members heretofore not enumerated, corresponding improvement by the churches in their own methods of keeping membership lists would tend to reduce the apparent membership by dropping the names of members who have been lost track of. As a matter of fact, the churches, or some of them, have been pruning their membership lists during these sixteen years. One of the large denomi- nations dropped from its rolls in a single year, if my memory serves me, something like 50,000 names — the number was certainly very large. Such pruning of membership lists is likely to offset any apparent increase that may result from changes in the methods of the census.

But it is not clear that improved census methods tend to increase the apparent number of members. On the contrary, the universal desire of the denominations to make a large showing might be favored by the looser rhethods and checked by the more cautious ones. Whether the improve- ment in census methods has, in fact, had any influence, I cannot say from any data in my possession. But, on the whole, in view of the known pruning of membership lists within denominations, I am inclined to think that the apparent membership increase shown by the new census is a real one."

In reply to a question whether the increase in membership appears in the cities as well as the country at large. Professor Coe said in substance :

"Separate statistics for cities are not given in Bulletin 103. But we can secure an approximate answer to the question by considering the distribution by states. Thus, the increase of membership in proportion to population is high in the North Atlantic states, with their great aggrega- tions of laborers in manufacturing centers, but low in the South Atlantic states, in which the population is more scattered. But there is increase in both. In general, the Catholic membership has increased with the greatest