84 Michael Hout
the American Baptist Convention, and the National Baptist Convention. In all, the GSS
codes 177 Protestant denominations, the distinction between Roman Catholic and
Orthodox Christianity, three Jewish denominations, five non-Christian faiths, and no
religion.
Very few other surveys take religion that seriously. Researchers affiliated with the
Gallup Polls, most notably George Gallup, Jr., have written extensively about religion.
But the Gallup data are much harder to use because of design changes, wording changes,
and few attempts to enumerate more finely than seven or eight Protestant categories.
Just as an example, the ubiquitous question about Americans’ belief in God at first
appears to be an important time series stretching back to the 1930s. Two important
wording changes break that trend line at crucial points; most recently the addition of
the phrase “or a higher power” to the question in 1976 reversed a downward trend in
response to the simpler question “Do you believe in God?” (see Bishop 1999).
CONCLUSION
Demography and religion have a fruitful past and a promising future. We can claim
Durkheim’sSuicide(1897/1951) as the first study in over a century of research linking
demography and religion. Researchers have looked at the consequences of religion for
demography – first in the fertility studies from the 1930s to the 1980s, more recently
in studies of religion and longevity – and (less often) at the consequences of demogra-
phy for religion. Both kinds of research have illuminated social change and helped us
understand religion’s role in American society.
The future is not guaranteed. The cutting edge of this kind of research depends on
infusions of mass data. With no questions about religion in the census, the continuation
of long-term studies such as the GSS are essential to our ability to keep doing this
important work.