Scores on the two indexes of religious diversity which serve as key variables in the empirical analyses of the cities in this study are listed below. The technique for computing these scores is illustrated in detail by Lieberson (1969; cf. Agresti and Agresti, 1977; Gibbs and Poston, 1975; Teachman, 1980). A short summary of the process will suffice here.

A city's score on either index is calculated by squaring, successively, the relative proportions of church membership reported for each relevant religious group, summing these quantities across all such groups represented in the city, and subtracting this total from one. The calculation is complicated in one minor way: the total number of members of “Other” religious bodies is first divided arbitrarily into three segments, of one-half, one-third, and one-sixth of the total, respectively, before squaring and deduction from one. This procedure, recommended by Lieberson (1969: 861), is designed to adjust the index to reflect the heterogeneous composition of the residual category and, accordingly, to prevent its size from affecting the diversity score in as direct a manner as do the sizes of the other, more descriptive, classifications. One exception to this practice was allowed: because Latter-day Saints (Mormons) were in the majority at both time points in their home base, Salt Lake City, their membership was extracted from the residual category and treated as a fourth religious subcommunity in the calculation of that city's overall level of religious diversity.