Demographic Methods for the Sociology of Religion 81
divorced, and so on. In religious research, this is likely to be more simple: A person
cannot convert from religion A to B unless she is an A to begin with. The most basic
activity in demographic research consists of measuring “rates” – the ratio of the number
of events to the number of people at risk of having an event. Most people are familiar
with the idea of a fertility rate, defined as the ratio of births to women of childbearing
age. Similarly, the marriage rate is the ratio of the number of marriages to the num-
ber of unmarried people; the divorce rate is the ratio of the number of divorces to the
number of married people. Rates are important because they estimate the probability
that the event in question will happen to an individual much more accurately than do
estimates that mix into the calculation people who are not at risk of having the event
occur.
All of this linking people to the risk of events comes together in a simple equation
that is true by definition: The number of events equals the probability that an event
will occur to a person at risk of the event times the number of people at risk. This simple
reexpression of the obvious becomes important when change occurs. The number of
events may change over time if either the rate or the number of people at risk changes.
So, for example, the number of births in the United States rose from 1980 to 1989 even
though the birth rate did not because the number of women between fifteen and forty-
nine years old increased. This is useful because while probabilities refer to behavior of
individuals, the number of people at risk is the factor that refers only to the population
and does not involve behavior per se. When a change can be attributed to a change
in the probability of an event occurring, then the explanation lies in something that
influences the behavior of interest. By contrast, if the number of events increases or
decreases because the number of people at risk changed, then “demography” is the full
explanation – as in nobody behaved any differently, there just happened to be more
people to act in the usual way. When demography is the full explanation, theories
about behavioral change are irrelevant.
The most obvious application in the sociology of religion would be to note that
the number of church members in a given locale or denomination rose because the
population increased. Trivial as it sounds, this was an important point to be made
when the Archdiocese of San Francisco closed several parishes in commercial districts
while opening new suburban parishes. TheSan Francisco Examinerasked in an editorial
why the residents of the commercial districts were giving up religion. The newspaper
missed the point that as the office buildings replaced apartments, the population in
those districts declined. The people still there were as religious as ever – they used that
fact about themselves to lobby the bishop to reverse his decision. But there were fewer
of them in the old neighborhood and more in the suburbs. The reallocation of priests
made demographic sense and told nothing about the relative piety of downtown and
suburban Catholics.
HETEROGENEITY AND EXPLANATION
The idea of linking people at risk of events to the rate at which those events occur has
even greater payoff when the rates in question vary systematically across important
categories. Then the distribution of the population across those categories can come
into the explanation of observed changes in either the number of events or in the overall
rate at which those events occur. Most characteristically, the mortality rate varies a great