Change in worker/nonworker ratios for women.

Effects of age, birth year, and observation year on work behavior are estimated by minimum logit chi-square regression, using data from Current Population Surveys. Most of the observation-year effect is captured by indicators of the business cycle. Further, a model which permits the cycle effects to vary by age is preferred over any additive model. The ratio for young women is most depressed relative to that for older women when times are "good".