As an addenda to a question about citations (see "Descriptors in a citation and a question about Church Censuses"), Ann Gilchrist wrote:
"From what I have been able to determine these books where created to meet a request from the US Census Bureau to develop statistics of churches in the US. I have located one Census Bulletin <http://www2.census.gov/prod2/decennial/documents/1890d4-01.pdf > dated 23 July 1891 which states, "This bulletin. which is the third devoted to statistics of churches,..." Does anyone know when the other 2 bulletins where complied and where I might find them?"
At least four approaches come to mind for pursuing these bulletins or the data therein:
1. (The simplest) Check Google Books, Archives.org, Cornell Library’s Digital Collections, and similar sites that are noted for OCR'ing old works.
2. Analyzing the bulletin that appears at this URL, we see that it is identified as “Bulletin No. 101.” If the bulletin is the first of three, as it seems to be, then we might hypothesize that the next two parts of the series might appear as Bulletins Nos. 102 and 103. On that supposition, we might alter the “1890-d4-01” portion of the URL to see whether “-02” and “-03” might turn up the next two items. That test gave the following results: “1890-d4-02” yielded Bulletin 106. “1890-d4-03” yielded Bulletin 110. “1890-d4-04” yielded Bulletin 114. “1890-d4-05” yielded Bulletin 120. The pattern suggests that the Census Bureau may have an incomplete set of those bulletins.
3. The statistics are likely drawn from the 1890 decennial census. Bearing in mind that the Census Bureau, starting in 1850, created a “statistical compendium” from the results of each decennial census, and that statistics on churches regularly appear in these compendiums, we might hypothesize that the information in these interim bulletins was subsequently reproduced in the social-statistics volumes prepared for the 1890 census. Many of the Census Bureau's statistical compendiums have been digitized for online access. A file list at www2.census.gov/prod2/statcomp/documents/ provides the links. The volumes we spotchecked for the 1890s-era, dealt with commercial matters rather than social issues. It is possible that, with more exploration, you may find other similar sites online.
4. Do you have access to a university library with a robust Government Documents Department? If so, you are likely to find there the “social statistics” volumes that would have the rest of the religious statistics you seek. Your state's major "research university" and your state's "A & M" university would typically have the most complete sets of these compendiums.