The two sets of times series data examined are data for the years 1941-57 and 1958-78. The earlier set of data only includes city data, while the later set encompasses both urban and nonurban areas. For the 1941-57 data, the imprisonment rate generally shows more declines than rises, and to a limited extent, the crime index rate rose more than it declined. The data from 1958-78 exhibit a much greater degree of statistical variation than the first set. Analysis reveals that incarceration rates declined steadily from 1961 to 1972, and then began a steady rise, reaching an estimated 143 per 100,000 in 1978. In contrast, the crime index rate rose fairly consistently throughout this period. Thus, the unlagged correlations between the crime and imprisonment rates for either data set are not statistically significant, but one of six lagged correlations from 1958-78 is significant, as are four of six from 1941-57. The inconsistency in correlation proves little guidance for the development of correctional policy. Considering these findings, William Nagel's support for a moratorium on prison construction seems to be a reasonable, and perhaps even conservative, reading of available policy and management data, rather than a radical proposition for change. The article concludes that although there are short-term pragmatic advantages to building more prisons, the lack of empirical support for the idea that imprisonment has social benefits suggests that vast resources should not be put into construction of new prisons. Three tables and nine footnotes are included. (Author abstract modified)