Repeat abortion.

A reanalysis of the repeat abortion experience of New York City residents during July 1, 1970 to June 30, 1972 is undertaken on the basis of a probability model that generates repeat abortion ratios as a function of assumptions about fecundity, contraceptive efficiency, and exposure lengths. Tested are three hypotheses put forward by Daily et al. in a 1973 analysis: (i) the low repeat abortion ratio of .0245 is attributable in part to underreporting of registered induced abortions as repeat ones; (ii) a major part of the rise in repeat abortion ratios, from virtually zero to six percent over four consecutive six-month intervals, is explainable in terms of the rising volume of exposure time to risk of repeat abortion relative to the stream of initial abortions; and (iii) the higher abortion ratios of women in their twenties compared to those of older or younger women is ascribable to "differences in fecundity and intercourse frequency." Support is found for the first two hypotheses, and a mixed outcome for the third.